JP4072193B2 - Method and apparatus for measuring relative viscosity or relative flow of a sample solution - Google Patents

Description

  The present invention relates to the field of viscosity measurement. In particular, the present invention relates to the field of capillary viscometers, which discloses a novel three-capillary capillary viscometer used to measure the viscosity of a solution. The present invention also relates to the field of digital signal processing, and more particularly, the present invention relates to viscosity and relative applications that can be applied directly in viscosity detectors or viscometers used in chromatography to measure fluid viscosity. It relates to a method of measuring the general flow.

  The measurement of fluid viscosity is a vital part of liquid chromatography (LC). In particular, in Size Exclusion Chromatography (SEC), a viscosity detector in combination with a concentration detector provides important information about the molecular weight in polymer analysis.

  The most basic capillary type viscometer is a single capillary viscometer. This capillary viscometer is based on the idea that as the fluid passes through the capillary, the pressure drop across the capillary is proportional to the viscosity and flow of the fluid. According to Poiseuille's law,

Are in a relationship.
In practice, the pressure transducer is placed across the capillary to measure the drop in pressure. If the flow is constant, the pressure is only proportional to the viscosity of the solution. Viscometers are very simple, but have the disadvantage that the pressure output is also proportional to the solution flow rate. Thus, a small disturbance in the flow of fluid in the capillary will cause a corresponding small pressure drop, or even a pressure drop greater than that caused by the viscosity of the solution.

The flow in the capillary basically depends on the flow of the pump device and the temperature change of the entire device. All of the slow and fast flows created by the pump are clearly detected by the viscometer. Also, any temperature change in the chromatograph causes expansion and contraction of the solution, which in turn causes flow disturbances that are also detected by the viscometer. Even a change in viscosity due to the injected sample can potentially cause flow disturbances. The turbulence in the viscometer is also caused when the pump device cannot react quickly to the fluctuating pressure load caused by the sample passing through the viscometer.

  These problems all impose very stringent requirements on the pump and the entire device in order to obtain good performance from a single capillary viscometer. The pumping device must provide a very character and constant flow that is unaffected by either slow drift or fast displacement normally associated with pump reciprocation. In order to eliminate flow errors due to temperature changes, the entire apparatus must be maintained at a constant temperature. To prevent flow disturbances due to sample viscosity, the volume in the apparatus and the size of the tube must be carefully considered.

  There are several types of capillary viscometers. The single capillary viscometer described above was originally described in US Pat. No. 3,837,217. Another single capillary structure is shown in US Pat. No. 4,286,457. A single capillary structure has a number of drawbacks, including sensitivity to temperature, sensitivity to small pump pulsations, and general effects on small device turbulence such as infusion.

  A structure consisting of multiple capillaries is described in U.S. Pat. No. 4,463,598 (Haney). Heyney discloses a bridge-type viscosity measuring device having two separate branches. Each branch has two capillaries arranged in series. These branches are connected at the top and bottom by a common inlet and outlet line. A pressure transducer for a bridge with a dead-end connects the two bridges across the two bridges in the middle between the first and second capillaries, with two branches at these points. Measure the pressure change at both ends. In normal operation (viscosity in both branches is the same), there is a fluid flow flowing down both branches without the same resistance and pressure differential. In operation, fluids of different viscosities are introduced into one branch and when it enters the capillary, a pressure difference begins to occur and reaches a maximum when the sample enters the entire capillary. This pressure difference is then measured by a transducer and the relative viscosity of the two solutions is determined by mathematical processing. The disadvantage of this structure is that the capillaries need to be balanced so that the resistances are approximately equal.

  U.S. Pat. Nos. 4,627,271 and 4,578,990 issued to Scott D. Abbott and Wallace W. Yau (referred to as Abbott et al.) Describes a differential pressure capillary viscometer that can be used to measure viscosity without being affected by flow rate and temperature pulsations. These patents include a solute injection valve upstream of two capillaries separated by a large deposition column used to trap the solute from the reservoir into the device, and the solvent only Discloses a viscometer adapted to flow in a second capillary. The change in pressure across both capillaries is measured and converted into an electrical signal, which is fed to a differential logarithmic amplifier. The output signal of the differential logarithmic amplifier is correlated to the natural logarithm of relative viscosity. Both intrinsic viscosity and intrinsic viscosity are mathematically correlated with experimentally measured values for relative viscosity. Although the devices disclosed in these patents provide viscosity measurements that are unaffected by flow rate and temperature pulsations, this device is capable of high-speed flow transients or high frequencies such as those caused by pump devices or sample injectors. Sensitive to the flow pulse.

U.S. Pat. Nos. 4,793,174 and 4,876,882 to Yau show that a large deposition column has been removed and instead a small separation volume is placed between two capillaries. A similar device is disclosed, except that This device effectively eliminates the delay associated with large deposition columns while remaining unaffected by flow rate and temperature pulsations. However, the devices described in these patents require closely matched capillaries to obtain the best performance and they are sensitive to fast flow transients and high frequency flow pulses.

  Further, all of the above structures use a pressure transducer with a “dead-end” connection that must be drained for proper operation. This device also creates additional problems as described below. Thus, all of these structures have some drawbacks, none of which has shown strong and correct operation to date. Thus, there is a need for an improved structure for a capillary viscometer.

  The present invention is an apparatus for measuring the viscosity of a sample solution, and is provided with an inlet pipe for conveying a sample solution flow Q, in fluid communication with the inlet pipe, and provided downstream of the inlet pipe. A fluid diverter (which divides the flow Q into flows Q1 and Q2), a first capillary R1 disposed in the flow Q1 downstream of the flow divider, and a flow Q1 downstream of the capillary R1. The first delay volume D1, the second delay volume D2 disposed in the flow Q2 downstream of the flow divider, and the second capillary R2 disposed downstream of the delay volume D2 in the flow Q2. A first flow-through type conversion having a third capillary R3 located downstream of the retarding volume D1 in the flow Q1 and a hydraulic connection located in the flows Q1 and Q2. Vessel T1 (this connection measures the pressure difference across the capillary R1 And a transducer T1 generates a signal proportional to the pressure difference across the capillary R1) and a second through-flow transducer T2 having this hydraulic connection located in the flow Q2 (this connection Is arranged to measure the pressure difference across the capillary R2, and the transducer T2 generates a signal proportional to the pressure difference across the capillary R2), and the signals from the transducers T1 and T2 Signal processing means for processing to generate relative viscosity information for the sample solution and relative flow information for the sample solution flow.

It is an object of the present invention to separately provide relative viscosity and relative flow information.
Another object of the present invention is that the solution always flows through all of its components without having any fluid closed end and without requiring drainage requirements like other viscometer structures. It is to provide a once-through viscometer.

Yet another object of the present invention is to provide a relative viscosity output that is unaffected by both the low and high frequency components of the viscometer flow.
Yet another object of the present invention is to provide a capillary viscometer that does not require matching or balancing of capillary limits, and therefore the relative viscosity output is not affected by capillary limit tolerances.

Yet another object of the present invention is to provide a method for compensating for non-linearities in the flow versus pressure relationship.
Yet another object of the present invention is to provide a method for compensating for the effects of pressure transducers having various fluid properties.

  Furthermore, according to the present invention, the problem of fast flow rate turbulence is solved by a digital signal processing method relating to the dynamics of the two pressure measurements usually included in viscometer structures consisting of a plurality of capillaries. This method can be used with any viscometer having two or more capillaries and at least two pressure transducers. Using this technique, the output of the viscometer detector is not affected by both the low and high frequency components of the viscometer flow. The performance of the viscometer is significantly enhanced without the use of special pumping equipment or any flow smoothing device.

The present invention provides several critical improvements to viscosity detection that are solved only by the present invention. The main advantages of the present invention are shown in the following description. Relative Viscosity Output and Relative Flow Output The viscometer of the present invention provides relative viscosity information and relative flow information separately. Relative viscosity (η _rel ) is a function over time of the ratio of the solution viscosity (η) to the solvent viscosity (η ₀ ) while the solution is flowing through the viscometer. The relative flow (Q _rel ) is a function of the time of the ratio of the solution flow (Q) to the solvent flow (Q ₀ ) at a specific “reference” time.

Relative viscosity information is determined according to the following equations: “specific viscosity” (η _sp ), “inherent viscosity” (η _inh ), “reduced viscosity” (η _red ) and “η _red ” There is a direct application in size exclusion chromatography (SEC) as a means to determine “Intrinsic Viscosity” (η _intr ).

In the formula, C is the concentration of the polymer solution.
Relative flow Relative flow information can also be used in size exclusion chromatography to correct retention time pulsations due to flow limitations. This provides better retention time repeatability between successive sample injections, which is a fundamental requirement in this analytical technique. Retention time correction can be applied not only to relative viscosity chromatography but also to other chromatography from other detectors connected in series with the viscometer. In addition to this, the relative flow signal is a variable diagnostic means for determining the correction function of the pump device.

  The methods and apparatus for measuring viscosity known in the prior art are partially or completely independent of the viscometer flow, but lack the ability to provide flow related values. Therefore, the prior art must be used with a pump device that provides a constant flow to ensure repeatability of the retention time through a series of samples. With respect to relative flow values, the viscometer of the present invention can tolerate errors in the supplied flow in a series of samples.

Flow- through The present invention is a flow-through type in which the solution always flows through all of the components without the closed end of any fluid and without the discharge requirements required for other viscometer structures. Viscometer. In prior art viscometers that use various pressure transducers, including viscometers consisting of a single capillary, the solvent in the transducer is stagnant and this solvent acts only as a pressure transmission medium. Each side of the transducer is connected to a connecting tube with a through path and a “T” type coupling.

Not only does the solvent in the transducer have to be drained all the time to remove contaminants, but the format is changed every time to maintain good pressure transfer of the solvent in the transducer. This discharge can be done manually or automatically by means of a solenoid valve. Manual drainage is a safety issue arising from the normally used harmful solvents. Automatic draining has the price and reliability issues due to the valves and controls required.

In the present invention, since the solvent always flows through the transducer cavity, the discharge of the transducer is not required and the transducer is always in the best operating condition.
Viscosity Output Insensitive to Fast Flow Changes Prior art viscometers provide viscosity output values that are not affected by low frequency flow components such as flow drift and slow flow pulsations. However, these prior art viscometers are sensitive to high-frequency flow components such as high-speed flow pulsations, such as transients due to pumping device pulses or chromatographic sample injection. These prior art viscometer constructions require a smooth pumping device or additional equipment to reduce or filter these high-speed turbulence. The viscometer of the present invention provides a relative viscosity output that is unaffected by both the low and high frequency components of the viscometer flow. Thus, it is not necessary to use a specific pumping device or any flow smoothing device.

Unlike some prior art viscometers that require capillary independence of capillaries , the present invention does not require capillary alignment or balance in any way. Thus, the relative viscosity output is not affected by the tolerance of capillary restriction. This facilitates the manufacture of the viscometer since there is no need for any kind of capillary preparation or viscometer fluid balance.

  Furthermore, there is no requirement for relative restriction of the capillaries. Although some capillary combinations have advantages over others, these capillaries need not be uniform and need not be of any special feature size. Also, the differential pressure transducer used need not be the same size. These can be of different full scale pressures. The only requirement for the capillary is that the pressure drop across the capillary must be within the dynamic range of the pressure transducer. The pressure drop must not be so great that the transducer saturates and the signal to noise ratio should not be unacceptably small.

  Also, as will be described in the detailed description below, only one of the capillaries only requires special consideration from a chromatographic point of view (internal volume and shear rate). Only one capillary must be ensured by a specific inner diameter and length that matches the intended internal volume and shear rate. Other capillaries do not require such a requirement, so the limitations of these capillaries can be obtained with various inner diameters and lengths. This further eases the design considerations of the capillary.

Preferred Embodiment The preferred embodiment of the present invention uses three capillaries, two retarding volumes, and two differential pressure transducers. These components are described in detail below.

1. Description of Concept FIG. 1 is a configuration diagram of a once-through viscometer composed of three capillaries. From this figure, three capillaries (R ₁ , R ₂ , R ₃ ) and two delay volumes (D ₁ , D ₂ ) can be recognized relative to it.

  A capillary is a tube member (usually made of stainless steel) of a length and an inner diameter that causes a drop in pressure as the solution flows through it. As mentioned above, the pressure drop is given by Poiseuille's law. In FIG. 1, the capillary is represented by a sawtooth waveform similar to the electrical resistance symbol.

The retarding volume is also a tubular member, but has an inner diameter that is much larger than the capillaries, which results in a negligible pressure drop as the solution flows through it. The purpose of the delay volume is to delay the arrival time of the distribution (effect) peak (SEC peak) on the capillary, and this delay time is proportional to the internal volume of the delay volume and inversely proportional to the flow of the solution. These delay volumes are sized so that the peak of the distribution is delayed by a time at least equal to the width of the peak.

The solution flow (Q) is divided into two branches. The branch part Q ₁ has two capillaries R ₁ and R ₃ , a delay volume D ₁ is arranged between these capillaries, and the other branch part Q ₂ is the same as the delay volume D _2. Another capillary R ₂ is provided downstream of the delay volume D ₂ . The pressures P ₁ , P ₂ and P ₃ are measured between any two of the three capillaries by a pressure transducer. However, the third capillary must always be present in order to obtain the aforementioned flow-through function or the intended viscosity output. The arrangement of the pressure transducer is described in detail below.

Initially, a certain solvent flows through both branches Q ₁ , Q ₂ of the viscometer. Thus, the three capillaries and the two retarding volumes are filled only with the solvent (viscosity η ₀ ). This is the baseline state.

When a distribution peak (viscosity η) is introduced into the viscometer, this peak splits into branches Q ₁ and Q ₂ and enters capillary R ₁ , but does not enter capillary R ₂ or R ₃ . R ₂ and R ₃ remain filled only with solvent (viscosity η ₀ ) due to the delay volume. Then, because the viscosity of the peak of the distribution is different from the viscosity of the solvent, the flow limitations of capillary R ₁ will vary, but capillaries R ₂ and R ₃ will still be filled with solvent so that their flow limitations. Does not change. However, as the flow restriction changes, the flow split into the two branches changes, so the pressure drop across all three capillaries changes. The differential pressure transducer measures two of these pressure drops and can determine relative viscosity and relative flow from these measurements.

When the distribution peak passes completely through R ₁ and into the retarding volumes D ₁ and D ₂ , there will be solvent (viscosity η ₀ ) inside all three capillaries and the viscometer will again be at baseline. It becomes a state. At this point, the viscometer provides a complete chromatogram of relative viscosity and forms the analysis part of the overall chromatogram. The remaining portion of the relative viscosity chromatogram and the effluent portion (described below) have no analytical significance.

When the distribution peak (viscosity η) leaves the retarding volumes D ₁ and D ₂ and enters R ₂ and R ₃ , R ₁ is filled only with solvent (viscosity η ₀ ), so the relative viscosity is of opposite polarity Peak (usually a negative peak). However, this peak is wider than the peak of the distribution due to the action of widening the peak due to the delay volume. As already mentioned, this part of the chromatogram does not have the following analytical significance.

When the distribution peak passes completely through R ₂ and R ₃ , the entire viscometer is filled with only the solvent (viscosity η ₀ ) and the viscometer returns to the baseline state. At this point, the viscometer can accept the new distribution peak.

Since only R ₁ is in contact with the distribution peak in the analytical portion of the chromatogram, only R ₁ needs special chromatographic considerations regarding length and inner diameter. R ₁ must be selected to produce the intended pressure drop while meeting chromatographic requirements for internal volume and shear rate.

During the analytical portion of the chromatogram, R ₂ and R ₃ are filled with solvent only, so they do not require special chromatographic considerations regarding length and inner diameter. These can be selected to produce the intended pressure drop regardless of the inner diameter and length.

If the delay volume D ₂ is also omitted, although viscometer is operated, it is impossible to obtain ideal performance. First, during the analysis of the chromatogram, the viscometer equation shown below is not valid because R ₂ is filled by the distribution peak (viscosity η). Even if another set of equations is established to explain this, one of the pressure transducers must always be placed before R ₂ (this is explained below). So the exact viscosity observed by R ₂ will be different from the viscosity recognized at R ₁ . Such a transducer connection causes a broadening of the width of the distribution peak in R ₂ , so that a viscosity equal to the viscosity in R ₁ cannot be guaranteed in R ₂ or is easily related to the viscosity in R ₁ . Even the attached viscosity is not guaranteed. In addition to this, R ₂ has a requirement for internal volume, which means that the length and inner diameter flexibility that exists with existing D ₂ is lost.

When choosing the size of the retarding volumes, these volumes must be large enough so that the distribution peak does not fall within R ₂ or R ₃ until the peak has completely passed through R ₁ . As described above, once the peak exits the delay volume and enters R ₂ and R ₃ , the analysis of the chromatogram is over and the rest of the chromatogram (the effluent portion) has no analytical significance. However, sufficient time has passed between successive samples to ensure that the previous distribution peak passes completely through R ₂ and R ₃ before the distribution peak of the next sample enters R _1. Must.

  Although not strictly necessary, in order for the distribution peaks to leave both delay volumes at the same time, these delay volumes must have a volume proportional to the flow therein.

In the specific case where R ₂ = R ₁ + R ₃ , the flow is split in half and both retarding volumes are equalized. This is very convenient but not essential.
In practice, differential pressure transducers are commonly used in all viscometer configurations to measure the slight pressure drop across the capillary of a transducer flow-through viscometer. These differential pressure transducers have two cavities of relatively large volume separated by a membrane. The pressure difference between both cavities distorts the film and the film distortion is converted to an electrical signal by magnetic coupling or other known means. One of the cavities is connected to one end of the capillary and the other cavity is connected to the other end of the capillary. Thus, the electrical output signal of the transducer is proportional to the pressure drop across the capillary. Each cavity has an inlet port and an outlet port. FIG. 2a is a conceptual diagram of this type of differential pressure transducer, in which the two halves of the transducer containing the cavity are labeled "+" and "-". . These signs mean that the transducer provides a positive signal if the positive cavity pressure is greater than the negative cavity pressure, and vice versa.

Transducer in “closed end” or standard connection state FIG. 2 b is a block diagram of a transducer used in a standard connection state that is typically not used by the present invention. The inlet port connects the cavity to the measurement location (both ends of the capillary tube) via a tube member and a “T” type bond. The discharge port is used to fill the cavity with the solvent used, and thus there is good pressure transfer between the measurement point and the membrane. Once the cavity is filled with solvent, the discharge port is closed and remains closed until a new discharge is required. Thus, in standard connection conditions, the transducer cavity is filled with a stable solvent, which transmits the pressure from the measuring point. In order to ensure correct operation of the transducer, it is necessary to discharge and open the device. The evacuation operation must be done all the time to maintain fresh and bubble free solvent in the cavity and must be ensured each time the solvent type is changed. If the same solvent is held in the cavity for an extended period of time, bubbles will form inside, which may cause noise in the pressure signal. If the solvent in the cavity is different from the solvent passing through the capillary, there is a potential ooze of solvent in the cavity that covers the flow of solvent in the capillary. This may cause noise or drift in the signal along with other physical or chemical problems with the solvent mixture. This draining can be done manually or automatically by a solenoid valve or other means. Manual drainage creates a safety issue due to the harmful solvents normally used in size exclusion chromatography. Automatic draining creates cost and reliability issues that result from the required valves and controls.

In a flow-through connection of a transducer in a flow-through connection, the discharge port of the transducer is never closed. Instead, the outlet port is used as an outlet port to “flow through” the solvent through the cavity. Both ports in each cavity can be used as inlet or outlet ports as long as one is the inlet and the other is the outlet.

  FIG. 2c is a block diagram of a transducer used in a once-through connection. Fluid flow enters the transducer cavity from one of the inlet ports. This flow leaves the cavity from the exit port and passes through the capillary. This flow then passes through the other cavity in a similar manner. Therefore, there is no closed end in the fluid flow. The differential pressure measured by the transducer is the same as the pressure measured in the standard connection state of FIG. This is true if the transducer port or cavity itself does not cause a significant pressure drop, and this is usually true because the port inner diameter and cavity volume are much larger than the capillaries. When this is not the case, it is always possible to design the converter to meet this requirement.

  This structure has significant advantages. In a once-through connection, no discharge of the transducer is necessary. This eliminates all emissions, performance, safety, cost and reliability issues that affect standard connections. The transducer is in ideal characteristics because there is no closed end volume to extract other solvents, and the solvent in the cavity is always as fresh and free of bubbles as it passes through the capillary. At this point, the transducer is being drained permanently. Since no discharge is required, there are no safety issues with respect to solvent handling in manual discharge operations. Similarly, a valve for automatic draining is not required and the cost and reliability issues associated with the valve are eliminated.

Transducer with a through-flow connection without peak broadening As described above, the through-flow connection of FIG. 2c and the standard connection of FIG. 2b provide the same pressure measurement at the transducer. This means that when the solvent viscosity is measured, the volume of the transducer cavity in front of the capillary “R” (positive cavity half) affects the measurement. The volume of the cavity is usually large enough to cause a broadening of the distribution peak width. Thus, the pressure peak measured at the once-through connection may differ from the value measured at the standard connection. However, the volume of the cavity behind the capillary (the negative cavity half) does not affect the pressure reading. A once-through connection that solves this problem is described below.

FIG. 2d is a block diagram illustrating a transducer that is used in a flow-through connection but does not cause peak broadening. A portion of the flow is diverted towards another capillary “kR” (usually k> 1), and the transducer cavities causing the broadening of the peak are connected by this new flow path. . The pressure measurement still relies only on the flow through “R”, but there is no cavity volume in the flow path before “R” and there is no broadening in the pressure peak. However, due to flow splitting, the measured pressure is smaller than that obtained with a standard connection and is reduced to k (k + 1). This is usually a problem, but there is always the possibility that both capillaries can be increased or converted to obtain the same measurement as with a standard connection. The scale of the vessel can be transformed.

  Also, splitting the flow indirectly affects the measurement of the pressure peak because the amount of flow split while the peak passes through both capillaries is different. In the present invention, this is fully taken into account and a through-flow connection is used. In any case, if k >> 1, this effect can be very small and the delay volume D is inserted before the capillary “kR”.

  It is important to note that in this once-through connection, the pressure measurement is not affected at all by the volume of the transducer cavity, even if it is very large. This substantially alleviates the mechanical requirements of the transducer which are essentially reduced until subject to low flow limitations. Figures 6b and 6c illustrate this embodiment.

In the present invention illustrated in flow-through connection diagram 1 in the present invention, it is preferable to use a flow-through connection shown in FIG. 2c for a capillary R ₂ and R _3. Due to the delay volume, these capillaries are filled with solvent (viscosity η ₀ ) during the analytical part of the chromatogram. Thus, there are no width-expansion issues to consider in R ₂ and R ₃ and the connection of FIG. 2c can be used.

However, in the case of capillary R ₁ (FIG. 1), the capillary R ₁ is problematic because it is filled with a distribution peak (viscosity η) in the analysis portion of the chromatogram. If the connection type in FIG. 2c is used for capillary R ₁ is spread problems width. Thus, for capillary R _1, as described below, it is necessary to use the connection portion of FIG. 2d.

The connection of the transducer for measuring P ₁ at both ends of the capillary R ₁ (FIG. 1) can be made by several arrangements that give ideal results. The pressure before R ₁ (in branch Q ₁ ) is the same as the pressure before delay volume D ₂ (in branch Q ₂ ). Since D ₂ is negligible as compared to the restrictions or capillary does not cause even the limit of any flow, the pressure of the previous R ₁ is also the same as the pressure after the D _2. Thus, the positive cavity of the transducer for measuring P ₁ can be connected to either side of D ₂ . Connection type in both cases is shown in Figure 2d, it does not cause broadening of distribution peak flowing through the R _1. Similarly, the negative cavity of the transducer measuring P ₁ can be connected to either side of D ₁ .

FIGS. 3a to 3d show four transducers for measuring the pressure across the capillary R ₁ (P ₁ ) without causing any width broadening of the distribution peak flowing through R ₁ with a flow-through connected transducer. Shows possible ways.

A preferred connection for measuring P ₁ is that shown in FIGS. 3a and 3d. This is because, if those which can not be ignored in the flow due to the delay volume limitation is a slight influence on the P ₁ in all cases other than the connection of Figure 3a. In the connection configuration of FIG. 3d, if the restriction in the delay volume is proportional to each flow, the effects of both delay volumes are canceled out in the connection of FIG. 3d.

FIGS. 4a to 4f show the topology of a transducer that measures the pressure (P ₂ ) across R ₂ without causing any widening of the distribution peak passing through R ₁ by means of a through-flow transducer. 6 shows six possible ways to form FIG. 4a shows a preferred connection (similar to that in FIG. 2c), and others show other embodiments based on a method for connecting the transducer cavities at various points of the same pressure.

FIGS. 5a to 5f show a transducer topology for measuring the pressure (P ₃ ) across R ₃ without causing any widening of the distribution peak passing through R ₁ by means of a through-flow connected transducer. 6 shows six possible ways to form FIG. 4a shows a preferred connection (similar to that in FIG. 2c), and others show other embodiments based on a method for connecting the transducer cavities at various points of the same pressure.

The present invention requires two of these three pressure measurements. Combining all possible methods for measuring P ₁ , P ₂ and P ₃ described above results in several possible through-flow connections of the two transducers used. In the case where P ₁ and P ₂ are selected pressure measurements, there are a total of 24 transducer through-flow combinations from FIGS. Also, 12 of these 24 combinations have one cavity for each transducer connected to the same location. This allows another 12 distinct combinations to be formed by changing the position where the transducer cavity is initially connected. Thus, there are a total of 36 transducer flow-through combinations using P ₁ and P ₂ .

If P ₁ and P ₃ are the selected measured pressures, there are a total of 24 transducer through-flow combinations from FIGS. 3 and 5. Also, 12 of these 24 combinations have one cavity for each transducer connected to the same location. This allows another 12 distinct combinations to be formed by changing the position where the transducer cavity is initially connected. Thus, there are a total of 36 transducer flow-through combinations using P ₁ and P ₃ .

If P ₂ and P ₃ are the selected measured pressures, there are a total of 36 transducer through-flow combinations from FIGS. 4 and 5. Also, 12 of these 36 combinations have one cavity for each transducer connected to the same location. This allows another 12 distinct combinations to be formed by changing the position where the transducer cavity is initially connected. Thus, there are a total of 48 transducer flow-through combinations using P ₂ and P ₃ .

  As a result of these combinations, there are a total of 120 possible separate inventive embodiments. All of these provide a relative viscosity measurement by flow through without causing a broadening in the distribution peak. Figures 6a to 6c are block diagrams of three separate embodiments or examples. Figures 7a to 7c show three similar embodiments using a "closed end" or transducer in standard conditions.

Preferred Embodiment FIG. 8 illustrates a preferred embodiment of the present invention. The main design consideration for practical experiments of the present invention is that only three capillaries cause significant flow restrictions. All other elements used in this structure produce a negligible pressure drop compared to that due to capillaries. The following description refers to the elements labeled in this figure.

The solution flow (Q) reaches the viscometer via a small inner diameter tube (1) to minimize the broadening of the peak width. This flow is divided into _two components Q ₁ and Q ₂ in a “T” type coupling (2) with a small internal volume.

Q ₁ passes through the following elements: That is, the capillary R ₁ (3) in which the pressure drop P ₁ occurs, the union (4) having a large inner diameter, the connecting pipe (5) having a large inner diameter, the delay volume D ₁ (6), the connecting pipe (7) having a large inner diameter, negative cavity of the transducer _{T 1} (8), a large connecting pipe inner diameter (9), the inner diameter of the large union (10), capillary _R 3 (11), a large coupling inner diameter (12) and the large connection tube having an inner diameter ( 13)

Q ₂ is to pass through the following elements. That is, the inner diameter is large connecting pipe (14), delay volume _D 2 (15), having a larger inner diameter connecting pipe (16), transducer _{T 1} of the positive cavity (8), a large connection tube having an inner diameter (17), positive cavity of the transducer _{T 2} (23), a large connection tube having an inner diameter (18), capillary _R 2 (20) of the inner diameter of the large union (19), the pressure drop _{P 2} occurs, the inner diameter is large union (21), large connection tube having an inner diameter (22), passes through the negative cavity of the transducer _{T 2} (23) and having a larger inner diameter connecting pipe (24).

The two flow components Q ₁ and Q ₂ are reconnected in a “T” type union (25) with a large inner diameter, and then the entire flow Q exits the viscometer via a tube (26). The tube 26 or any other element connected to the rear of the viscometer does not affect the performance of the viscometer unless the maximum absolute pressure of the transducer is exceeded. Further, the “T” type union (25) and the outlet pipe (26) are unnecessary when the two branches are ventilated to atmospheric pressure.

The electrical outputs of the converters T ₁ (8) and T ₂ (23) are sent to an analog / digital converter (27). The digital output from the transducer is processed in a digital signal processor (28) to obtain relative viscosity and relative flow information.

  Capillaries are generally made of stainless steel tubing with an inner diameter in the range of 0.2286 millimeters (0.009 inches) to 0.3556 millimeters (0.0014 inches). The preferred size is 0.012 inches. Other materials such as aluminum or industrial plastics can also be used. The connecting pipe and the union having the large inner diameter generally have an inner diameter of 1.016 millimeters (0.040 inches) or more. The retarding volume is typically made of tubing having an inner diameter of 1.5748 millimeters (0.062 inches) or greater.

Capillary R ₁ is the only capillary that contacts the distribution peak during the analysis of the chromatogram. Thus, only one capillary needs specific chromatographic considerations regarding length and inner diameter. R ₁ is the dynamic range of the transducer while meeting the chromatographic requirements of internal volume (typically 8-16 microliters) and fluid shear rate (typically 3000-3000 / sec). Selected to produce the intended pressure drop within. Conversely, R ₂ and R ₃ do not require special chromatographic considerations regarding length and inner diameter because they are filled with solvent during chromatogram analysis. These can be selected to produce the intended pressure drop regardless of the inner diameter and length.

The flow path of the distribution peak in SEC in the present invention is as follows. Initially, a constant solvent flow (Q ₀ ) passes through the viscometer. Thus, all viscometer components are filled only with solvent (viscosity η ₀ ). This is the baseline condition, in which the transducer shows a reading of the reference pressures _P1Baseline and _P2Baseline .

The solution flow (Q) containing the distribution peak (viscosity η) reaches the viscometer through the tube (1). In the “T” type union (2), the peak is divided into two parts (not necessarily equal), part of the peak goes to capillary R ₁ (3) and the other part is tube (14 ) To the delay volume D ₂ (15). This has two effects. That is, the pressure drop across the capillary R ₁ (3) is increased by a higher viscosity solution passing through the capillary (assuming η> η ₀ ), so that the fraction of the divided flow is in the baseline state It changes for the divided flow.

Some of the peaks passing through Q ₂ (15) take some time to exit this delay volume. Within this time, a portion of the peak Q ₁ passes through the capillary R ₁ (3) and enters the retarding volume D ₁ (16) through the union (4) and the tube (5). Similarly, this part of the peak takes some time to leave the delay volume D ₁ (6).

While the peak (viscosity η) passes through the capillary R ₁ (3) and is delayed by the retarding volumes D ₁ (6) and D ₂ (15), the remaining components of the viscometer are still solvent (viscosity η ₀ ). Thus, the change in pressure drop measured by transducer T ₂ (23) relative to the pressure measured during baseline conditions is directly related to the change in flow split. However, the change in pressure drop measured by transducer T ₁ (8) relative to the pressure measured in the baseline condition is related to both the change in viscosity and the change in the flow split ratio. From the two pressure measurements, the relative viscosity can be derived in the signal processing portion as described below.

Once the peak has completely passed through R ₁ (3) and into the delay volumes D ₁ (6) and D ₂ (15), the “analytical” portion of the chromatogram is complete and all relationships Some relative viscosity information is obtained. The remaining part of the viscometer component, including R ₁ (3), is filled with solvent (viscosity η ₀ ). In this situation, the viscometer is in the baseline state. The peak then exits the delay volume D ₁ (6) and becomes the “flush” portion of the chromatogram. The peak will come out of both retarding volumes at the same time or several times depending on its size and flow split ratio, since this part of the chromatogram has no particular significance for the analysis. It does not matter.

A broader peak passes from the retarding volume D ₁ (6) through the negative cavity of the tube (17) and the transducer T ₁ (8). This peak then enters the capillary R ₃ (11) through the tube (9) and the union (10). This results in a split flow change that is measured as a pressure drop by both transducers T ₁ (8) and T ₂ (23). The peak then passes through joint (12), tube (13) and "T" type union (25), where the flow from both branches is reconnected and through tube (26). get out. A broader peak emerges from the retarding volume D ₂ (15), and the positive cavity of tube (16), transducer T ₁ (8), tube (17) and transducer T ₂ (23) positive. Pass through the cavity. This peak then enters the capillary R ₂ (20) through the tube (18) and the union (19). The change in viscosity at R ₂ (20) causes a change in shunt as measured by transducer T ₂ (23) as a change in pressure drop. Peak, then Union (21), the tube (22), a negative cavity of the transducer _T 2 (23), the tube (24), the flow from both branches is connected again "T" type of Union (25) and exit pipe (26). While the solvent (viscosity η ₀ ) is passing through capillary R ₁ (3), the combined effect of the peaks (viscosity η) passing through capillaries R ₂ (20) and R ₃ (11) is relative viscosity. Is a peak of the opposite polarity (usually a negative peak) to that of a relative viscosity distribution peak (usually a positive peak). However, the negative peak is wider than the broadening due to the delay volume and the distribution peak due to dispersion.

Once the peak passes completely through the capillaries R ₂ and R ₃ , the entire viscometer is again filled only with solvent (viscosity η ₀ ). This returns the viscometer to baseline and is ready to accept a new distribution peak.

Signal processing Transducer Signal Processing The two characteristics of viscometer pressure transducers are primarily responsive to the sensitivity to high velocity flow changes in a multi-capillary capillary viscometer. One is some non-linearity of the transducer response and the other is their dynamic response.

  In some cases, the pressure transducer signal does not exactly follow Poiseuille's law, so the relative relationship between the measured pressure and the viscometer is somewhat non-linear. Although other members also act like bent capillaries due to space limitations of practical viscometer construction, pressure transducers themselves can respond to most of this non-linearity. is there.

  Some non-linearity in the transducer response usually affects the low frequency response of a multicapillary viscometer because the change in pressure normally contained in the transducer is very small compared to the full scale of the transducer. Does not reach. However, if there are large pulsations of flow, whether low or high, they are slightly distorted by the non-linear response. Eventually, normal signal processing by the viscometer of the transducer signal cannot cancel the flow turbulence.

  In practice, the difference from the nonlinearity of the transducer determines the magnitude of this effect. If the non-linearity of each transducer response is exactly equal, processing of the transducer signal by the viscometer completely cancels the flow turbulence.

  Another characteristic of the viscometer pressure transducer already mentioned is the dynamic response of the pressure transducer. As in the case of non-linearity, it is the change in the dynamic response of the transducer that affects the sensitivity of the multicapillary viscometer to high speed flow changes.

  Due to several factors, the dynamic response to high flow changes of transducers used in multiple capillary viscometer structures is different for each transducer. This changes the “shape” of the pressure transient measured at each transducer. The main elements that can respond to this phenomenon are the various arrangements of transducers in the viscometer fluid arrangement, the mechanical capacitance of each transducer membrane and the capacitance of the fluid, and the structure used. The fluid capacitance of the delay volume. The digital signal processing method disclosed in the present invention is essentially to solve the problem of dynamic response in two stages: non-linearity and multicapillary device as described below. This method does not depend on the type of structure of the multi-capillary viscometer and is therefore applicable to any viscometer with two or more capillaries and at least two pressure transducers. This results in a significant performance increase because the viscometer can eliminate high frequency flow disturbances without affecting the kinetic response.

  These two stages of the method are referred to in the following as “Pressure linearization” and “Dynamic equalization”, which are usually performed in this order. However, these can be done in the reverse order so that dynamic equalization is performed first and pressure linearization second, in which case the high-frequency flow through the viscometer The increase in perturbation rejection performance may not be as good as in the normal sequence step.

  Figures 15 and 6b are block diagrams of two multi-capillary viscometer structures that can benefit from the method of the present invention. FIG. 15 shows a viscometer consisting of two capillaries and two transducers, and FIG. 6b shows a viscometer consisting of three capillaries and two transducers.

After the pressure signals are “preprocessed” using the method of the present invention, these signals are further processed using conventional processing corresponding to the structure of the particular form of viscometer. In the present embodiment, such a processing method is given in part B, which is
Subsequent to this, signal processing of the converter is performed.

  Importantly, the same results obtained using the digital signal processing method disclosed herein can be obtained using analog electronics at the output of the pressure transducer. While this analog solution is possible, it is difficult to implement and absorbing changing viscometer conditions is usually less adaptable.

  The two stages of the method are described below for the specific case of a viscometer structure using two pressure transducers. Similar concepts can be applied to other viscometer structures with two or more pressure measurements. In the following description, P1 and P2 represent pressure signals generated from both transducers.

1) Pressure linearization Each of the pressure transducers is linearized with respect to the flow, with the introduction shown more specifically below using a quadratic fit. This compensates for any minor non-linearities that may have a relative pressure to flow relationship and increases the independence of the relative viscosity to changes in viscometer flow.

  The basic idea of this correction is that for each pressure signal, a curve fit of the pressure relative to the flow is used to calculate a pressure value that is linearly related to the flow near the baseline. It is to be.

The following examples use a secondary fit. Pressure measurements in two known streams are made for each transducer and these measurements are used to calculate a secondary fit factor. While these pressure measurements are made, the viscometer must be filled with solvent only, so there should be no trace of old solvent or sample inside the viscometer. One of the pressure measurement _{(P1 Baseline} and _{P2 Baseline)} is performed at a flow of baseline viscometer has during normal operation _{(Flow Baseline).} Other pressure measurements (P1 _Mid and P2 _Mid ) are made at an intermediate flow point (Flow _Mid ) near the baseline flow. If pump pulsations appear in the pressure signal, special care must be taken while making these pressure measurements. In order for the pump pulsation to not affect the measurement, the measurement must be calculated as an average of instantaneous pressure readings during an integer number of pulse periods. The factor used in the secondary fit is given by:

With these coefficients, the linearized pressures (P1 _lin and P2 _lin ) are calculated as a function of the pressures (P1 and P2) measured sequentially using the following equation:

These linearized pressure values are used in the following steps. 2) Dynamic equalization The purpose of this stage is to equalize the dynamic response of both pressure signals. This creates a relative viscosity signal that is independent of high frequency flow components such as pulses from the pump device or fast transients from the sample injection device.

  Due to the filtering action formed by the pressure transducer itself as well as the various fluid components in the viscometer, the flow transients in the viscometer are represented as filtered transients in the pressure transducer. Since this filtering action is different for each pressure signal, the filtered pressure transient is also different.

  The result of this different dynamic phenomenon is that the flow transient in one of the pressure signals (eg P1) is “smoothed” compared to the same flow transient in another pressure signal (eg P2). This smoothing is only due to the viscometer. Viscometer fluid geometry studies generally assume that either pressure signal is a smoothed pressure signal. “Dynamic Equalization” applies the same smoothing to the P2 pressure signal as the viscometer hydrodynamics applies to the P1 pressure signal. “Dynamic equalization” does not change the P1 pressure signal anyway. The end result is that both pressure signals respond equally to flow transients, and these transients usually correspond to the relative viscosity signal being calculated or to a specific type of multicapillary viscometer structure. These transients are offset during the processing of the pressure signal.

  In any case, kinetic equalization does not transform the steady-state response of the viscometer that relies solely on signal processing corresponding to the viscometer structure, such as the relative viscosity calculation described in the next step. It is important to pay attention.

  “Dynamic equalization” finds what the transfer function smooths the P1 pressure signal compared to the P2 pressure signal, and then applies this transfer function to the P2 pressure signal. Although other methods can be used, this method is easy to perform on a sampled data device.

  The basic model of most multicapillary viscometer structures discloses that this transfer function has a pole zero pair in the Laplace transform plane. This translates into a function in a discontinuous time domain where the current output of the transfer function depends on the current input and depends on the previous input and output of the sampled data device according to the following equation:

Output (t) = A.Output (t-1) + B.Input (t) + C.Input (t-1)
In the formula, A, B and C are weight coefficients, “t” represents the current sample, and “t−1” represents the previous sample. A condition that must be satisfied in order for this transfer function not to change the steady state is that the sum of the three coefficients must be unity. _Knowing these three coefficients, an “equalized” P2 pressure signal (P2 _eqzd (t)) can be obtained from the P2 pressure signal (P2 _lin (t)) linearized by the following equation:

  To obtain a more accurate representation of the transfer function, more coefficients and samples of “t−2”, “t−3”... “T−n” can be used. The next process is the same as described below, but generally three coefficients are appropriate. The calculation of the coefficients A, B and C is a number of processes based on a series of pressure data points collected during the baseline flow. Another possible way to calculate the coefficients from the actual data is to use a learning neural network, and those skilled in the art will find other methods as well.

2.1) A row of pressures A row of linearized P1 and linearized P2 data points is collected by a viscometer filled with solvent only, and such a row of data is referred to as “baseline flow”. The viscometer will have during normal operation. This sequence must be long enough to contain at least one flow transient. If there are pump pulses, it is sufficient to include a few pulses in the train. In the absence of pump pulses, these can be generated artificially by slightly changing the flow a known number of times. Their purpose is to capture the response of both pressure transducers in the same flow transient, from which the coefficients can be calculated.

  It is fundamental to this calculation that two columns are filled with samples of pressure collected at approximately the same time (each element in both columns corresponds to a sample taken at the same time).

In the following description, the variable “t” indicates one specific element of the column, the variable “t−1” indicates one element before the element “t”, and “t−2”. Indicates the two elements before the element “t”. It is assumed that the sampling rate at which the data points were collected will be the same as the sampling rate used during normal data collection. It is also hypothesized that the degree of element increase means increasing the time that elements are collected. This column is named “column P1” and “column P2”, and each includes elements “P1 _lin (t)” and “P2 _lin (t)”.

2.2) Normalization At this stage, the two columns are normalized to 1 to eliminate the steady state component of the pressure data point and leave only the variable component. The average value of the data points is calculated according to the same method as described in the “pressure linearization” stage to obtain the baseline pressure values (P1 _Baseline and P2 _Baseline ) for each column. Each pressure data point in the column is then divided by its respective baseline pressure value. The standardized columns are “column N1” and “column N2”, each containing elements “N1 (t)” and “N2 (t)”.

2.3) Filter Noise present in the pressure signal greatly affects the calculation of the coefficients B and C. Thus, the signal in the standardized column is filtered and improved by a single pole filter. The filtered columns are “column F1” and “column F2”, which contain elements “F1 (t)” and “F2 (t)”, respectively.

  The coefficients in the above equation represent a specific filtering amount for a specific sampling rate. The value of the two coefficients in each case depends on the sampling rate used and the amount of noise present in the pressure signal, but their summation must always be unity.

2.4) Calculation of “Row A” and “Row B” Since the two filtered rows have no steady state components, each element has the same “kinetics” as described above. It is related by the “equalization” equation. That is, the equation that gives P2 _eqzd (t) from the value of P2 _lin (t) is the same equation that provides the value of F1 (t) from the value of F2 (t). Therefore, F1 (t) and F1 (t-1) are expressed as follows.

From these equations, A and B can be expressed as a function of only the values of F1 and F2.

By applying all the elements of “column F1” and “F2” to these equations, a set of values of A and B, that is, “column A” and “column B” can be obtained.
2.5) Calculation of coefficients A and B In an ideal and complete device, all elements in "column A" should be the same and elements in "column B" should also be the same . However, this is not the case in real devices due to noise present in the pressure signal. Each element in the column is different, but there is a clear tendency towards a specific value that is the value of the coefficient to be determined. Therefore, the value of the coefficient A is the value that is most likely to occur between the elements of the “column A”. The same applies to “column B”.

  Before doing this step, some elements in both columns with irrational values are eliminated. Both coefficients A and B have limitations imposed by the model used to represent the viscometer dynamics and the sampling rate used. Normally these coefficients should be positive and less than or equal to 1, but for a particular viscometer structure and sampling rate, a narrower margin can be calculated. All elements in the row that have values that exceed this limit must be rejected by noise in the device.

One method for easily calculating the value most likely to occur during the rest of the elements in the column is described below, but other methods may work as well.
In order to calculate the coefficient A, an average value and an intermediate value of each element in the “column A” are calculated. Then, if this average value is greater than the intermediate value, all elements having a value greater than the average value are eliminated. If the average value is smaller than the intermediate value, all elements having a value smaller than the average value are excluded. New average and median values are then calculated and more column elements are eliminated according to a similar method. This method is repeated until the average value is equal to the intermediate value. This situation usually occurs when only two elements remain in the column. The coefficient A is a value obtained to make the average value and the intermediate value equal. A similar method is followed to calculate the coefficient B from "column B".

2.6) Calculation of coefficient C When the values of the coefficients A and B are determined, the coefficient C is calculated from the requirement that the sum of the three coefficients must be 1. Therefore, C = 1−A−B.

B. Transducer signal processing At this stage, the pre-processed pressure signal determined above is used to perform basic relative viscosity and relative flow calculations.
In FIG.

Due to Poiseuille's law, the pressure drop across the capillary is

Given by.
Also, the relative relationship between the flow in the two branches is

It is.
With these equations, the following equation depending on the flow “Q” can be obtained.

When the viscometer is filled with solvent only (η = η ₀ ) and there is a baseline flow Q ₀ in the viscometer, the above equation becomes:

For the preferred embodiment of FIG. 8, the two transducers used measure the pressure drop across capillaries R ₁ and R ₂ only. Therefore, only the equations P1, P2, P1 _Baseline and P2 _Baseline are taken into account.

  Using these four equations, the following equations for relative viscosity and relative flow can be derived.

  When the delay volume D2 is removed, the pressure drop across the capillary is

It becomes.
Using the same method as described above, the relative viscometer when D ₂ is not present and the expression of the relative flow can be obtained as follows.

The formula for relative viscosity is the same, but the formula for relative flow is different.
For those embodiments where the pressure drop across R ₂ and R ₃ only is measured and a retarding volume D ₂ is present, the relative viscometer and relative flow equations are as follows:

When the D ₂ has been removed, wherein the relative viscosity is the same, wherein the relative flow changes as shown below.

Finally, for an embodiment where the pressure drop across R ₁ and R ₃ only is measured and a retarding volume D ₂ is present, the equations for relative viscosity and relative flow are as follows:

When the D ₂ has been removed, but the formula of the relative viscosity is the same, wherein the relative flow changes as below.

Experimental Results FIGS. 9-14 show chromatograms obtained using the preferred embodiment of FIG. Analysis conditions other than those described above are as follows.

Figures 9, 10 and 11 show the chromatograms of relative viscosities obtained by standard solution injection.

The signal to noise ratio of FIG. 9 is 275, FIG. 10 is 2280, and FIG. The inverted “dips” in the chromatogram are due to the distribution peaks passing through R ₂ and R ₃ .

  In order to clearly demonstrate the advantages of relative viscosity and relative flow signals that are not affected by flow changes, the following experiments were performed. The solvent flow was reduced to 0.9 ml / min and a 2630 molecular weight solution was injected. During the 60 minute chromatogram, the flow increased linearly to 1 milliliter per minute and returned.

  FIG. 12 shows a relative flow chromatogram accurately tracking flow transients. FIG. 13 shows a chromatogram of relative viscosity whose amplitude is not actually affected by flow transients. However, peak retention times are clearly affected by flow transients and are clearly different from those in the chromatogram of FIG. 9 (about 2 minutes difference).

  Using the relative flow information in the real-time process or the advance process, the peak in FIG. 13 can be repositioned at the right time, so the retention time is as if the flow was constant at 1 microliter / minute. It is. FIG. 14 shows a corrected relative viscosity chromatogram, which has a peak retention time equal to the chromatogram of FIG.

  FIG. 16 shows the viscometer output of a multi-capillary viscometer structure without the pretreatment method of the present invention, and FIG. 17 shows the viscometer of the same viscometer structure with the pretreatment method of the present invention. Output is shown. The multi-capillary viscometer format used is a once-through viscometer consisting of three capillaries that generate a relative viscosity output. The test conditions are summarized in the following table.

  A dual piston AC pump with a fluid filter is connected directly to the viscometer, which is probably the worst possible condition for a capillary viscometer. FIG. 16 clearly illustrates the high speed flow transients of the pump without using the pretreatment method described herein. FIG. 17 uses a plot scale similar to FIG. 16 and shows that the transient due to pump pulsation is reduced by more than 50 times when the pretreatment method is used.

  Small disturbances still visible in the plot of FIG. 17 can be reduced using more than two factors in dynamic equalization, which further reduces pump pulsation. However, the three factors have been found to be sufficient for practical applications.

Equivalents Those skilled in the art will recognize equivalents to the specific embodiments of the invention specifically described herein or will recognize the invention using minor routine experimentation. Such equivalents are intended to be encompassed in the scope of the claims.

It is a block diagram of this invention which shows the relative connection of a capillary and a delay volume. a is a block diagram of a differential pressure transducer showing the arrangement of two transducer cavities, inlet ports and outlet ports.

b is a block diagram of a differential pressure transducer used in a “closed end” or standard connection.
c is a block diagram of a differential pressure transducer used in a once-through connection.

d is a block diagram of a differential pressure transducer used in a once-through connection with no broadening of the peak width.
It not a d is flow connected, _{R 1} 4 single possible to measure both ends pressure capillary _{R 1 (P 1)} by the converter without causing even the spread of any width to the peak of the distribution through the (distribution peak) It is a block diagram which shows a simple method. a through f show six possible ways of measuring the pressure (P ₂ ) across the capillary R ₂ by means of a transducer without any spread of the width of the distribution through R ₁ , connected through flow. It is a block diagram. a through f are connected through-flow and show six possible ways of measuring the pressure (P ₃ ) across capillary R ₃ by means of a transducer without causing any width broadening of the distribution peak through R1. FIG. a through c are block diagrams of three examples selected from the 120 possible embodiments of the present invention. FIGS. 6a-c are block diagrams of the three embodiments of FIGS. 6a-c, but showing the "closed end" or converter in a standard connection. It is a block diagram of preferable embodiment of this invention. It is a chromatogram of a standard relative viscosity with a narrow width of 2630 Mw. 12 is a standard relative viscosity chromatogram with a narrow width of 1260000 Mw. 9 is a wide standard relative viscosity chromatogram of NBS706. It is a chromatogram of the relative flow of a flow transient. FIG. 13 is a 2630 Mw narrow standard relative viscosity chromatogram with the same flow transients as FIG. 12. FIG. 13 is a relative viscosity chromatogram corrected for time using the relative flow chromatogram of FIG. It is a block diagram of the viscometer provided with two capillaries and two converters. FIG. 6 is a graph of a relative viscosity baseline in a once-through viscometer consisting of three capillaries without a pre-processing algorithm according to the present invention. FIG. 4 is a graph of a relative viscosity baseline in a once-through viscometer with three capillaries with a pre-processing algorithm according to the present invention.

Claims (2)

A method for obtaining a relative flow (Qrel) of a sample solution comprising a solute and a solvent, comprising:
A diverted flow downstream of the first capillary to communicate with the first capillary to form a first flow having a flow Q1 and to communicate with a second capillary to form a second flow having a flow Q2. Wherein the first capillary has a geometric resistance R1 and the second capillary has a geometric resistance R2 and is provided with an inlet tube in communication with the shunt And
A delay volume in communication with the first capillary for receiving the solution and causing a delay in fluid movement; and a geometric resistance R3 in communication with the delay volume for receiving fluid Further providing a third capillary having
Means for measuring the pressure drop P1 across the first capillary in the presence of the solution, and means for measuring the pressure drop P1 (baseline) across the first capillary in the presence of the solvent And means for measuring the pressure drop P2 across the second capillary in the presence of the solution, and measuring the pressure drop P2 (baseline) across the second capillary in the presence of the solvent. Means for measuring, a pressure drop P3 across the third capillary in the presence of the solution, and a pressure drop P3 across the third capillary in the presence of a solvent (baseline) Providing a means for measuring
The relative flow (Qrel) is

A method determined by one of the relative relationships. The method of claim 1 , comprising:
A second delay volume D2 is disposed in communication with the flow Q2 downstream of the flow divider and upstream of the second capillary;
The relative flow (Qrel) is

The method determined by one of the formulas.

Images