CA2231524A1 - Detection of near surface defects by reversed thermography - Google Patents

Description

DETF,CTION OF NEAR SURFACE DEFECTS BY REVERSED THERMOGRAPtIY
field .of the invention The present invention relates to a method for the detection of near surface defects by reversed thermography.
The invention also relates to an apparatus for carrying this method which is particularly useful for the inspection of thermal sprayed engine block cylinder wall linings.
Back around Thermal spraying is a generic teen that encompasses an array of processes used to apply coatings of various high melting point materials (metals, ceramics, etc.) as a spray of dropleas heated and propelled by some form of flame or plasma. The coating is built as "splatted" droplets solidify and accumulate on the substrate's surface.
Thermal spraying is extensively used in the aerospace industry and, more recently, has moved into the automobi le industry. in particular, for the manufacturing of engine blocks.
Presently, engine blocks are made of cast aluminum with steel sleeves lining the inner surface ofthe cylinders. The approximately 1/16" thick sleeves are inserted in the mold prior to the casting of the aluminum. ror several years now, auto makers, sometimes in collaboration with tlhermal spraying equipment manufacturers, have been developing coating processes that would replace the steel inserts. Significant savings in manufacturing costs as well as weight reduction and performance enhancement has fuelled this general trend throughout the industry.
Because it is very thin (.005" to .008"), the wear resistant coating (generally steel}
performs better because heat is extracted more efficiently than through a 1/16" thick sleeve.
However, bonding of the coating to the aluminum substrate is critical and bonding defects are generally considered unacceptable. Since an in-service breakdown of the coating leads to failure of the engine, the quality and reliability requirements for these processes are extremely high and it is uncertain if they can be met without some form of inspection method.
There is presently no mel.hod of detecting these defects that could reasonably be used in a production environment. The present Applicant, Tecnar Automation Ltee, became aware of thia situation through its contacas with the ford Motors Corporation on a related matter and the decision was made to pursue this opportunity and develop a suitable method.

2 fns~rtion bX thermo~ranhv The basic principle of tlrermographic inspection is to establish heat flow conditions through the inspected component and detect disruptions in the heat flow pattern caused by potential defects. Some form o~f infrared (IR) detection device (IR camera for example) measures the surface temperature of the component. Defects perpendicular to the direction of the heat flow will cause the most disruption and therefore will be the easiest to detect.
Since bonding defects in coatings are obviously parallel to the component surface, one must establish heat flow conditions perpendicular to the component surface in order to maximize contrasts caused by these defects. A typical setup will use some form of heat source (generally a laser) to uniformly heat an area of the component surface. If there is no defect, the temperature of the area is uniform. Any defect that restricts the flow of heat from the surface toward the inside of the component will appear as a hot spot within the heated area. figure identified as "prior art" illustrates such a setup.
Of course, the mechanism used to heat the surface must not itself generate IR
radiation that would reflect on the inspented area and blind the IR detector. A laser is the most convenient from that perspective because its precise color can easily be filtered out of the IR
detection device. However, the cost per watt of power can be prohibitive.
rurthermore safety and limited mean time between failure are also major concerns. Other heat sources have been used such as joule heating with an electrical current or even a flow of hot air.
ummarv of the invention The present invention is based on the discovery that, since engine blocks come out of the coating station at a temperature of about 90°C, one could actually use reversed thermography to inspect the coating on cylinder walls.
Thus, the invention provides a method and an apparatus for the detection of near surface defects in components such an engine block cylinder wall, comprising the basic steps of applying a flow of cold air to am area to be inspected, thereby establishing a heat flow from the inside-out, and subjecting thie cooled area to an inspection by reversed thermography in order to locate any cold spots which would be indicative of defects.
As it can be understood, the amount of heat extracted by the cold air is a function of the difference in temperature between the air and the surface. In steady-state conditions, that temperature reaches an equilibrium where the cooling rate equals the amount of heat

3 continuously flowing from the hot mass underneath the cooler surface. Where there is a bonding defect, the equilibrium temperature will be lower because heat flow towards the surface is disrupted. Therefore, defects will appear as cold spots, in reversed contrast with traditional thermography.
Brief descri~ltion of the drawings Figure 1 identified as "prior art" is a schematic representation of the way bonding defectt> can be detected by laser thermography;
Figure 2 is a schematic representation of the way bonding defects can be detected by the meahod according to the invention, using air cooled reversed thermography;
Figure 3 is an IR camera image of a cylinder surface with defects while cooled with compressed air;
Figure 4 is a curve giving the temperature profile along a line running across two defects; and Figure 5 is a schematic representation of an apparatus according to the invention for use to inspect engine block cylinders.
Detailie de cri~tion of the inven_ lion The way bonding defects can be detected by the method according of the invention is illustrated in Figure 2. As already indicated hereinabove, this method is not restricted to the detection of defects in cylinder walls. As a matter of fact, it can detect any defect whose size, shape and orientation significantly disrupts the heat flowing perpendicularly to the surface of a component.
If such is needed, cold components can be heated prior to inspection. All that is required is that the component be massive enough to act as a "virtually infinite" heat sink (or, in this case, "cold" sink) for as Kong as it takes to establish steady state heat flow conditions through the inspected layer. In other words, the component has to be much thicker than the inspected layer.
In carrying out the method according to the invention, it is assumed that heat flows in only one direction (inside-out). C~f course, heat also flows sideways and can go around a defect, blurring the contrast at the edge of deeper ones. Therefore, the resolution of the method is

4 limitedi by the depth of potential defects, i.e. the thickness of the inspected layer.
Another factor that contributes to the efficiency of the method according to the invention is the thermal conductivity of the material. 1f it is high, a physical discontinuity will cause a much greater disruption in the heat flow l.han if the material is itself a good thermal insulator. Therefore, contrasts in thermal barrier coatings will be lower than in the case of cylinder walls where the steel coating is a fairly good heat conductor and aluminum is an excellent one.
Figure 3 shows images of artificial bonding defects in a test cylinder wall.
The aluminum cylinder was 3" in diameter and 5" long. Wall thickness was 3/8" with a .010" thick steel coating on the inside. For convenience, the cylinder was longitudinally cut in two.
An IR camera was aimed l:oward the inside surface of one of the half cylinders which was then heated with a heat gun (a kind of"industrial" hair dryer) to about 60°C. The picture from tlhe IR camera showed a uni~Formly warm surface.
Compressed, room temperature air was then blown from a nozzle directly onto an area at the center of the camera's field of view where defects were known to be present. As can be seen in Figure 3, the cone shaped area struck by the compressed air appears as shaded in the IR
image because it is understandably cooler. Near the center, two defects are clearly visible as cold spots within the continuously cooled area. The shape and location of these spots are in perfect agreement with the defect map given by the test cylinder maker.
Figure 4 gives the temperature profile along the faint line marker visible in Figure 3.
As can be seen, the defects are 4 to 6°C cooler than the surrounding area. Considering that the temperature difference between tLie sample and the compressed air is about 50°C, 4 to 6°C is a very significant contrast.
Insne~etion of cylinders In the simple experiment described above, the geometrical complexity of inspecting the inside of a 360° cylindrical surface was circumvented by cutting it in half, thus allowing direct., perpendicular viewing of tine interior. For real engine blocks, an optical arrangement has to be designed to allow viewing from the top, such as illustrated by way of example only in Figure 5.
As can be seen, a simple mirror allows the IR camera located on top of the block to inspect an area inside the cylinder. Compressed air should of course be blown onto the same area. The mirror and air nozzle; are located inside a cylindrical housing that shrouds the remainder of the cylinder's inner surface to prevent 1R radiation from that surface to blind the IR camera. The whole assembly pivots 360° to scan a complete cylindrical section of the surface. The head is moved up and the process is repeated until the whole cylinder has been

5 covered.
As can be appreciated, the whole process can be automated. A computer can easily collate; the various scan results to display an unfolded, continuous image of the whole cylinder which can be processed by image analysis software to signal an operator if the area and density (i.e. intensity of the contrast) of defects exceeds preset thresholds.
Of course, modifications could easily be made to the embodiment illustrated in Figure 5 withiout departing from the scope of the present invention.

Claims