ZA200200884B

ZA200200884B - Fluxing agent for dry application.

Info

Publication number: ZA200200884B
Application number: ZA200200884A
Authority: ZA
Inventors: Hans-Walter Swidersky; Alfred Ottmann; Heinz-Joachim Belt
Original assignee: Solvay Fluor & Derivate
Priority date: 1999-10-25
Filing date: 2002-01-31
Publication date: 2003-01-31
Also published as: KR100734411B1; KR20020021679A; JP5241463B2; JP2009061507A; DE50003674D1; UA71644C2; TW523437B; MY124744A; DE10049315A1

Description

The invention relates to a flux which can be used for dry application, and to the use thereof as a soldering flux.

For many years, it has been known to solder together components of aluminium or aluminium alloys, in particular heat exchangers for the automobile industry, using fluxes based on alkali fluorocaluminate. In such a case, the flux is usually sprayed on to the heat exchangers as an aqueous suspension. In the presence of a solder or of a solder- forming precursor such as silicon powder or potassium fluorosilicate, a stable, non-corrosive compound is formed upon heating the components to a temperature above the melting point of the flux. A process has indeed already been disclosed by DE-0S 197 49 042 with which the waste water produced in this procedure can be recirculated. However, other process parameters are critical: the concentration of the flux slurry must be monitored, the heat exchangers must be dried before heating, the flux slurries, which are likewise recirculated, may pick up impurities. These disadvantages can be avoided if the flux is applied dry to the components to be joined. This is the case in the dry-flux process. Therein, the dry flux powder is applied to the components electrostatically. The advantage is that no slurries need to be produced, that the concentration of the slurry does not need to be monitored, that it is not necessary to provide a separate drying stage for the components, and that no waste water is produced.

It is an object of the present invention to devise a flux based on alkali fluoroaluminate which can be conveyed

NC a 2 well pneumatically, can be sprayed well in the dry state and which adheres to the sprayed components well and is therefore suitable for the method of dry application (dry fluxing).

This object is achieved by the flux set forth in the claims.

The invention is based on the finding that the particle . size or the grain-size distribution of the alkali fluorocaluminate fluxes has an influence on the pneumatic transport, the spraying ability and the adhesive power of the flux particles on the components. It was discovered that it is advantageous if smaller and larger particles are contained in the flux and the ratio thereof is subject to certain rules.

The flux based on alkali fluorocaluminate according to the invention which can be used for dry application (dry fluxing) is characterised in that the volume distribution of the particles lies substantially within Curves 1 and 2 of

Figure 10. The particle-size distribution was determined by laser diffraction.

In a preferred flux, the volume distribution of the particles lies substantially within Curves 1 and 2 of Figure 11.

Figure 10 shows the lower limit (Curve 1) and the upper limit (Curve 2) for volume distribution curves of usable powders in the spirit of the present invention. This is the volume distribution in % c¢f the powders, cumulated, plotted against the particle size. Flux powders, the cumulative volume distribution of which lies on or within Curves 1 and 2 of Figure 10, are powders within the spirit of the invention.

The cumulative volume distribution of Curves 1 and 2 of

Figure 10 plotted against particle size 1s set forth in Table

A belcw.

RA h 3

Table A: Cumulative volume distribution plotted against particle size of Curves 1 and 2 of Figure 10

Q3 [%] Lower limit Q3 [%] Upper limit 0.45 0.25 3.00 0.55 1.40 4.00 0.65 2.00 5.30 0.75 2.70 6.80 0.90 3.80 8.80 1.10 5.00 12.20 1.30 5.80 15.80 1.55 7.00 20.00 1.85 8.50 25.00 2.15 10.00 29.00 2.50 11.50 32.50 3.00 14.00 41.00 3.75 17.00 53.00 4.50 16.00 63.00 5.25 19.00 71.00 6.25 23.00 79.00 7.50 28.00 86.00 9.00 33.00 90.00 10.50 38.00 94.00 12.50 40.00 96.00 15.00 42.00 98.00 18.00 44.00 98.70 21.50 48.00 99.50 25.50 54.00 100.00 30.50 65.00 100.00 36.50 77.50 100.00 43.50 89.00 100.00 51.50 93.00 100.00 61.50 94.00 100.00 73.50 95.80 100.00 87.50 96.00 100.00

Lower limit = Curve 1 Upper limit = Curve 2

Selection example: 40% of the volume is made up of particles having a diameter of 12.5 um or less.

It was established that fluxes having a cumulative volume distribution on or within Curves 1 and 2 of Figure 11 have particularly advantageous dry flux properties. Table B shows the numerical values of the cumulative volume distribution plotted against the particle size of Curves 1 and 2 of Figure 11.

™ 1 , 1 4

Table B: Cumulative volume distribution of the particle size of Curves 1 and 2 of Figure 11 am Q3 [%] Lower limit Q3 [%] Upper limit 0.45 0.94 2.28 0.55 1.53 3.49 0.65 2.19 4.73 0.75 2.91 6.00 0.90 3.91 8.07 1.10 4.97 11.69 1.30 5.89 15.30 1.55 7.03 19.58 1.85 8.43 24.20 2.15 8.91 28.19 2.50 11.76 32.18 3.00 14.58 37.01 3.75 18.94 43.07 4.50 22.24 48.09 5.25 25.31 52.30 6.25 29.74 57.13 7.50 34.30 64.82 9.00 37.26 72.07 10.50 38.78 77.06 12.50 40.25 81.89 15.00 41.87 86.27 18.00 44.20 91.28 21.50 48.13 95.12 25.50 54.67 97.45 30.50 65.04 98.91 36.50 77.82 899.70 43.50 89.38 100.00 51.50 96.55 100.00 61.50 98.64 100.00 73.50 100.00 100.00 87.50 100.00 100.00

Lower limit = Curve 1 Upper limit = Curve 2

The material according to the invention can be obtained by sieving off undesirable grain fractions, by mixing - material with different grain-size distribution.

The spraying factor is preferably 25, preferably 35, in particular 45 or more, and the ratio Hg) ,jq:Hp determined thereby is at least 1.05. The upper limit for the spraying factor was 85, preferably 83.5. The determination of the spraying factor and the ratio of H¢j,;49 to Hy (height of the expanded powder relative to the non-expanded powder) is described further below.

a " ' 5

The material according to the invention is very highly suitable for use as a flux in the dry fluxing process. In that process, the powder is introduced by compressed air or nitrogen from the storage container into a "spraygun" and 1is electrostatically charged therein. The powder then leaves the spraying head of the spraygun and hits the components to be soldered. The components to be soldered are then soldered, optionally assembled, in a soldering furnace, usually under inert gas for nitrogen, or by torch soldering.

The powder according to the invention has application- related advantages compared with known fluxes. For example, it has very good flow behaviour. This is ascribed to the selected distribution of particle size. This good flow behaviour results in the tendency to clogging ("build-up") being reduced. The material can be electrically charged very well. The material adheres very well to the components to be soldered. The flow of material is very uniform.

The invention will be explained further with reference to the following examples, without limiting its scope.

Examples

I pati ft} . listribution:

System: Sympatec HELOS

Manufacturer: Sympatec GmbH, System-Partikel-Technik

Set-up:

Measuring apparatus for determining particle-size distributions of solids by means of laser diffraction.

The apparatus consists of the following components:

laser light source with beam formation means, measuring zone in which the particles to be measured interact with the laser light, an imaging lens which converts the angular distribution of the diffracted laser light into a location distribution on a photodetector, a multi-element photodetector with autofocus unit and subsequent electronics which digitise the measured intensity distribution.

The particle-size distribution is calculated by means of the software WINDOX. The principle is based on the evaluation of the measured intensity distribution of the diffraction pattern (according to Fraunhofer). In the present case HRLD (high-resolution laser diffraction). The particle size of non-spherical particles is reproduced as an equivalent diameter distribution of spheres of identical diffraction.

Before measurement, agglomerates have to be broken down into individual particles. The aerosol of the powder which is required for measurement is produced in a dispersing apparatus, in this case RODOS system. The uniform supply of the powder to the dispersing apparatus is effected by means of a vibrating conveyor (VIBRI).

Measuring range: 0.45...87.5 um

Evaluation: HRLD (version 3.3 Rel.l)

Density of the sample:

Setting: 1 g/cm?

Form factor: 1 complex refractive index m=n-ik; n=1; i=0

Evaluation: be is the particle diameter in um.

Q3 is the cumulative volume percentage of the particles up to the diameter listed. q3 is the density distribution for the particle diameter pd ao ’ 7 x10 is the particle diameter at which the cumulative volume percentage reaches 10%. c_opt 1s the optical concentration (aerosol density) which occurred upon measurement.

M1.3 and Sv were not used for evaluation.

Start i ; {als

Two powders consisting of potassium fluoroaluminate with different grain-size distributions were investigated in terms of their properties for dry fluxing. The powders are obtainable by sieving off undesirable grain fractions. The grain-size distribution (volume distribution) is compiled below in table form. The particle-size distribution of Powder 1 ("coarser" material) is shown visually in Figure 1, and of

Powder 2 ("finer" material) in Figure 2.

~ ) . . 8

Table 1: Volume distribution of Powder 1 x0/ym Q3/% x0/um Q3/% x0/gm Q3/% x0/um Q3/% oss | 340 | 215 | tet | 000 | seer | 3650 | 9044 090 | 741 | 375 | 2864 | 1500 | e158 | 6150 | 100.00 iss | 1305 | e2s | 4ses | 2550 | esos x10 = 1.14 um X50 = 7.35 um x90 = 19.44 um

Sv = 2.033 m?/cm3 Sm = 8132 cm?/g copt = 6.27%

Table 2: Volume distribution of Powder 2 x0/pm Q3/% x0/pm Q3/% x0/um Q3/% x0/pm Q3/% 055 | 613 | 215 | 4035 | 900 | e438 | 36.50 | 10000 oso | 1408 | a75 | este | 1500 | esse | e150 | 10000 155 | 284s | 625 | soos | 2550 | sees x10 = 0.72 um x50 = 2.71 pm x90 = 7.26 um

Sv = 3.6046 m2?/cm3 Sm = 14418 cm?/g copt = 6.74%

First of all, the fluidising ability and the flow ability of

Powders 1 and 2 and certain mixtures of both were investigated.

«0 ‘ S

Apparatus used and method of performance: 1 measuring apparatus for determining the powder fluidising ability and powder flow ability (Binks-Sames powder fluidity indicator AS 100 - 451 195) was constructed on a vibration unit (Fritsch L-24). The measuring apparatus had a fluidising cylinder with a porous membrane at the bottom. 250 g of the powder to be investigated in each case was introduced into the cylinder, the vibration unit was switched on and a uniform flow (controlled by a flow meter) of dry nitrogen was introduced through the porous membrane into the powder. The powder expanded; in order to adjust the equilibrium, the gas was allowed to act for 1 minute. The fluidising ability of the respective powder can be determined by measuring the height before and after expansion.

The fluidising ability and flow ability of the respective powder were determined by means of the so-called "spraying factor". The spraying factor is a combination of the expansion factor (fluidising ability) and the mass flow of the powder (flow ability). The spraying factor represents an important factor for dry flux application. It was determined as follows: as already described above, the powder to be investigated in each case was expanded in the fluidising cylinder. Then a hole formed in the side of the cylinder was opened for 30 seconds, and the powder leaving the cylinder through this hole was collected in a beaker and weighed. The ratio of the quantity of powder collected relative to the unit of time of 0.5 minutes is referred to hereinafter as the "spraying factor". In explanation, it should be mentioned that highly fluidisable, flowable powders have a spraying factor of 140. Very poorly expandable, poorly flowable powders have, for example, a spraying factor of 7. Table 3 below gives the spraying factors determined for pure Powder 1, pure Powder 2 and intermediate mixtures containing 90, 80, 70 ... 10% by weight of Powder 1, remainder Powder 2 to make up to 100% by weight.

Table 3: oo : Powder 1 = So Co hi Powder 2 ; Co . ‘Spraying factor’ in

Lk) RY 0 Agi0.S min). 100 0 71.88 90 10 63.56 80 20 35.54 70 30 25.33 60 40 22.51 50 50 21.52 40 60 14.76 30 70 13.83 20 80 11.28 90 9.77 0 100 7.35 1 Average of several measurements

In tests it was determined that good flow behaviour is obtained for a spraying factor of more than about 45 g/0.5 min.

The spraying factor can also be calculated as follows: a) The expansion factor is calculated (cm/cm):

Heiuiq:Ho with Hgy,iq = height of the expanded powder,

Hy = height of the non-fluidised powder, vibrator switched off and supply of nitrogen stopped.

The average of 5 measurements in each case of measuring points distributed over the diameter is determined. b) Flow of the powder in (g/0.5 min):

The weight of the powder which flows out of the hole in 0.5 min is determined as median value of 10 measurements.

Calculation of the median:

Median m = mg + m,/2 for 10 individual measurements with

Mg <M3<My <M; <MNg<My<My<Mg<Mj n<Mg

The spraying factor R, is then

R,(g/0.5 min) = m(g/0.5 min): expansion factor

Very surprisingly, the spraying factor did not change linearly with the composition of the powder mixture, but exhibited a strong jump in the properties in the range of about 80-90% of the proportion of Sample 1. This is shown graphically in Figure 3. The spraying factor is plotted in g/0.5 min against the percentage of the Powder 1 in the mixture. This proves that the content of fines in the powder has a great influence on the flow ability. ft) 1 1 Co as a function of the grain-gi ze distribution:

The adhesive power was tested by a very simple method which permits good conclusions to be drawn about the industrial usability of the powders investigated for dry fluxing.

A flat, square aluminium plate of dimensions 0.5 m x 0.5 m was electrostatically spray-coated on one side with the dry flux powder which was to be investigated. The loading with flux was weighed out; the plate was then dropped on to the ground in a vertical position from a height of 5 cm and the loss of flux was noted as a percentage of the original flux loading. 10 measurements were performed for each of the powders. Poorly adhering powders had a comparatively high weight loss compared with the low weight loss when using powders according to the invention (see Powder 3 and Powder practice:

Two different apparatus were used. One apparatus was a flux application apparatus ("fluxing booth") manufactured by

Nordson, suitable for semicontinuous application. Dimensions of the unit: 216 cm high, 143 cm wide, 270 cm deep. The most important components were a storage container, a spraygun, two filter cartridges and the control units. The component to be fluxed was placed on a grate which could be moved back and forth manually. The spraygun moved automatically from left to right and back again at intervals of about 21 seconds (21 seconds for 65 cm, i.e. the speed was 3.1 cm/sec.).

A container from ITW/Gema together with a spraygun and control unit was incorporated in this system as a second fluxing unit.

The distance between the spraying heads and the grate was 34 cm. incipl - FE

The Nordson container applied the principle of powder fluidisation in order to introduce the flux into the spraygun via a venturi pump and a feed hose. A stirring or shaking device in the container supported the fluidisation of the flux.

The ITW/Gema system had a container which had screw conveyors ("helix screw conveyors") for conveying the powder mechanically into a funnel. A venturi pump then conveyed the flux through a hose into the spraygun.

The ITW/Gema system was equipped with vibrators at some points in order to avoid clogging by the flux. The sprayguns operated at 100 kV for charging the powder.

The powders listed in the examples were used in the Nordson and ITW/Gema apparatus in order to investigate the uniformity of flux transport and the spraying operation and the loading of test pieces (heat exchangers having a surface area of 4.8 m2). First the control units were adjusted relative to the air throughflow and the screw speed such that flux loading of approximately 5 g/m? was achieved. Then the experiment was continued for 30 minutes without changing the setting of the apparatus. At intervals of 2-4 minutes, test pieces were placed on the grate for spraying with flux, and then were weighed out to determine the flux loading. Each test series comprised 10 or 11 measurements. The results are set forth in Table 4.

Nom es [es | or [ws |e | ie

Figures 4 to 7 show the flux loadings for the Nordson apparatus and the ITW/Gema apparatus plotted against time for

Powder 1 and Powder 2 in graph form. For Powder 2, the spraying head of the Nordson apparatus had to be blown free regularly in order to avoid clogging.

The 30-minute test investigations as described above were performed for additional powders. Powder 3 had the following properties: a measured value of Ry, of 59.25; Hfiuig:Ho (mm/mm)= 1.11; a loss of adhesion of 11.5%; and the following particle-size distribution: 90% of all particles had a size of <35.15 um; 50% of all particles had a size of <9.76 um, 10% of all particles had a size of <1.35 um. A peak maximum of the grain-size distribution was at 5 um, and the second- largest peak was at 20 um. The totalled volume distribution of this powder is shown in Figures 5 and 6 as an example of a 14 readily usable powder. This material yielded very good results both in the Nordson apparatus and in the ITW/Gema apparatus. "Spitting" was not observed in the apparatus, and nor was it necessary to blow off the spraying Lead. The layer produced was "very attractive". The flux coverage plotted against time is shown in Figure 8. A further material was

Powder 4, and it had a spraying factor of Ry, = 82.85;

Heyuiq:Hg was 1.10; the loss in the adhesion test was 16.7%; the particle-size distribution: 90% of all particles had a diameter of less than 28.6 pm; 50% of all particles had a diameter of 8.9 Hum; 10% of all particles had a diameter of less than 1.67 um; the grain-size distribution had a peak at 9.5 and at 20 um, and this material too yielded excellent results. Figure 9 shows the uniformity of the flux coating with Powder 4 on the heat exchanger plotted against time.

Acceptable results were also obtained with the following potassium fluorocaluminate Powder 5: R= 46.99; ratio Hgjyig:Hp = 1.05, loss coverage: 6.39%, particle-size distribution: 90% of all particles <19.84 um; 50% of all particles <7.7 pum; 10% of all particles <1.16 um, max. peak of the grain-size distribution at 13.

Claims

1. A flux based on alkali fluorcaluminate which can be used for dry application ("dry fluxing"), characterised in that the volume distribution of the particles lies substantially within Curves 1 and 2 of Figure 10.

2. A flux according to Claim 1, characterised in that the volume distribution of the particles lies substantially within Curves 1 and 2 of Figure 11.

3. A flux according to Claim 1, characterised in that it is a flux based on potassium fluorocaluminate.

4. A process for soldering aluminium or aluminium alloys, wherein a flux of Claims 1 to 3 is used which is applied dry and electrostatically charged to the components which are to be joined and the components are soldered with heating.