How to obtain the best system for inspecting rubber parts.

The first and most important step in obtaining the best system for inspecting rubber parts is to determine if the test you wish to perform is going to give you enough information to meet your requirements. For a check valve (also known as a duck bill), for example, does the part you are testing have a slit? Do you need to know if the slit is present; is the slit the correct size; is the slit in the correct direction; does the slit have small tears at the ends; does the slit allow enough flow; is the slit flow bi-directional; will the slit re-open at the correct pressure level; or will the slit hold the correct vacuum pressure? These are some of the considerations that need to be evaluated.

Another example is the pressure testing of a rubber-to-metal bonded gasket with raised sealing beads. This part needs to be tested by sealing the part between two plates, allowing the pressure to equalize, and then testing the presence of the molded sealing beads. There are always other factors that can destroy the ideal test. To begin, is the part flat; if the part is bent, how much is allowable; can the part have pieces of flash or injection sights flash folding over the testing area; is the test to be of the sealing beads or the entire molded edge; are we trying to find breaks or incomplete molding of the edges?

Rubber parts present a major problem to most inspection systems, due to the soft nature of the parts and the tendency of the parts to stick together; when a large number of parts is produced at one time, the parts are usually hard to orientate and organize. Most parts are black, which produces great shadow effects if you are checking diameters. This type of test is very easy to produce. But black parts make it very difficult to find surface irregularities, or cracks, or to find a slit in a part.

We will look at three parts which have been tested; the first with a flow test, the second with a pressure test and the third with an optical test. Altman Manufacturing can address all of the above situations and present solutions which will provide guarantees you and your customers demand.

Slit diaphragm 

This part was being slit with the use of a small blade like a scalpel or very sharp knife. The equipment included a rotary table producing a part every 10 to 12 seconds. The operation was loaded by hand through a light curtain and a palm button to start. The problem was the part needed to be more consistent and the inspection completed as quickly as the slitting operation. We changed the slitting operation to a shear type cutter with a supporting female.

With this slitter, we gave outstanding results and far longer service. Prior to the changes, the customer was stopping to change blades every two hours. Once the slitting system was changed to a shear type slitting system, the operation would run weeks at a time. Sharpening would require removing the rounding at the end of the shear type slitter punch or blade, and perhaps squaring the end of the support. The results were a far more consistent length of cut, less tapering of cut and less string pieces left in the slit. To confirm the presence of the slit in the final stage on the rotary table, we performed a flow test. We captured the part in a way that allowed the air to flow quickly through the part. Giving results in a second or less, therefore, the machine would now slit and test parts at a rate of 30 to 45 parts per minute.

Pan gasket

The testing of a large pan gasket presented a number of problems. Note that the part was presented to the testing system while it was still hot, but could be moved in and out without gloves. The part was checked with a laser beam to determine if the part was fiat. A minor amount of straightening was allowed by the operator. This gave the operator an incentive not to bend the part while removing the parts from the mold. The operator running the testing system also runs the molding press for single responsibility of quality parts. The system was controlled by a small process controller, which after determining that all inspections were correct, would then run the pressure test. If any test or gauge failed to meet the desired range, the process controller told the operator what needed to be adjusted and in the correct direction. Also note the use of an o-ring to reduce the area under pressure. This has a number of benefits. First, if you pressurized a whole pan gasket (about 10" x 24"), this would push open the test fixture with a gigantic amount of force. Second, the pressure would need to equalized between the force holding the plates closed and those pushing it apart. This takes time and slows the test. Third, all leads or wires must be sealed that pass through the test fixture. This is not easy when testing was to .01 of a psi (1/100 of one pound per square inch) (figure 1).

[FIGURE 1 OMITTED]

With the use of a small o-ring, the amount of press was reduced and the equalization was less than two seconds. The test was to pressurize the beads with an air regulator that was set to a 1/100 of a psi of the correct testing force. Then an air valve sealed off the test area and the pressure was held for three to four seconds. The length of time was adjustable, but in testing we found that three to four seconds were all that was needed to check the bead. The test could be adjusted to different pressures to see if only the sealing beads were broken or to see if the entire molded area was void. This test was needed due to parts that were not completely surrounded with the necessary molded sealing edge. The operator should have found the problem parts very easily, but due to the repetitive nature of the work, they all start to look very similar by the end of the day.

Baby nipple 

The last part we will consider is a baby nipple with two slits in the flange. This part needed to be located in the cavities prior to slitting, because the slits had to be in an exact location in the flange. The parts were presented to the machine by a two-cord conveyor. This allowed the parts to hang tip downward and gave the customer the option of using an operator or, in the future, changing to a feed bowl (figure 2). The part had two raised check valves that needed to be slit. These valves needed to be in the correct orientation prior to slitting. This was completed by using a laser light beam to detect the valves while the parts were rotated in a rotating cavity. At this point, the part was properly aligned, and the part could be loaded into the slitting cavity by a pick and place system. The parts were slit, and then the testing was needed to tell if the parts were slit. Note that this is where you, as a customer, must decide what you intend to inspect. Altman recommended a simple test to see if the blades went through the part by pressing down with a cylinder, and a vision system would detect the presence of the blades appearing through the rubber. Therefore, if the blades came through the part, the part was slit. And if the blade pushed the rubber up, the light would deflect or not show the same.

[FIGURE 2 OMITTED]

The customer later asked why we did not check to see that the part was in the correct rotation. The answer is that we could have if that was required, but it would have required a second camera and some additional costs. We also could have inspected the length of the slit, or how centered it was. Testing with a camera can be very quick, as long as light can "see" what you are trying to test. We have found that cameras need to be mounted very firmly and locked into position. Note that training is also needed to readjust the lens if it or the light source is moved. This is something most companies do not consider when ordering an optical inspection system. The most time in designing cameras is spent in getting the part to the camera and light source quickly. Once the part is presented to the cameras, as many inspections as needed should be made.