Bellows are among the most complex components that can be machined from PTFE (Teflon). Owing to the nature of PTFE, there exist several limitations to making the bellow by conventional, melt-processable means. However, the properties of PTFE make it invaluable in certain applications, requiring the bellows to be machined from solid PTFE.

For the most part, high-performance bellows are made using rubber or equivalent elastomers. The limitations of rubber include high temperatures and resistance to corrosion. It is likely that PTFE is only used in very specialised circumstances both because of cost and complexity.

The challenges for machining PTFE bellows include:

1. The bellow often needs to be machined out of a solid rod or bush. The process can result in wastage of up to 80% in some cases
    
2. The process of creating the convolutions needed to form the bellow requires special tooling and high-precision CNC machining. The smallest deviation in the machining process can result in a complete loss of the whole component. In cases where the final part is large, this represents a high risk of loss due to wastage and rejection
    
3. The raw material needed for making bellows needs to be of the highest purity and the final rod or bush that is moulded would need to be completely free from any micro-faults, such as cracks, inclusions or discoloration. For the best quality bellows, grades such as DuPont NXT or 3M TFM are used. More recently, Inoflon has come out with a resin grade – M690, which also works well in making bellows. These grades are modified grades, having a high purity and capable of more flexibility
    
4. The final produce needs to have total uniformity and be completely free from any irregularities in colour or dimension. Even the smallest deviation in wall thickness or the slightest micro-crack in the walls of the bellow can result in complete failure during operation.
    
5. The design of the bellow itself needs to accommodate the properties of PTFE. For instance, it may not be able to have a wall thickness that goes below 0.5mm. As a result, the maximum compressibility of the bellow is limited when compared with an elastomeric bellow. Not only are elastomers more elastic, but their melt processability allows for much thinner cross sections, meaning more compressibility

Despite these limitations, when applications require high corrosion or temperature resistance, the only option is to use PTFE (Teflon). Hence, designers work around the limitations of the material to create a bellow that would work in their conditions.

If properly manufactured, a PTFE (Teflon) bellow can withstand up to 2,000,000 cycles. Since most requirements call for anywhere between 500,000 and 1,000,000 cycles for the life of the bellow, this metric allows for a sufficient cushion.

Once manufactured, a bellow is subjected to a series of tests to confirm its long-term durability during application.

1. Burst pressure test
   The burst pressure of a PTFE bellows shall be at least four times the design pressure given by the manufacturer, after it has been subjected to 2000 cycles at 10 cycles per minute between its maximum axial extensions. The pressure and related temperature at which cycling is carried out shall be selected from the pressure/temperature graph supplied by the manufacturer. The pressure to produce failure shall be applied uniformly at such a rate that failure occurs within 5 minutes. As a minimum two tests shall be carried out; one at ambient temperature and the other at 180 °C.
    
2. Cycle testing
   No failure shall occur when a PTFE bellows is subjected to 100 000 cycles at 10 cycles per minute between its maximum axial extension, or a combination of axial and lateral extension, at a pressure and temperature selected from the pressure/temperature graph supplied by the manufacturer.
    
3. Temperature test
   The bellows shall be bolted to a mating flange and held at a temperature of 260 °C for 2 hours. After cooling, the PTFE flange face shall be examined to ensure no deformation or damage.
    
4. Dimensional inspection
   PTFE bellows shall be subjected to dimensional inspection, visual examination, liquid penetrant examination, holiday testing (if necessary) and a hydrostatic pressure test. A burst pressure test and a pressure shock test shall be applied if the bellows ordered are to be used for critical applications (this to be specified by the Principal).

Because of the extensive testing and care needed during manufacture, the cost of a PTFE (Teflon) Bellow does tend to be many times that of its elastomeric counterpart. As a result, these bellows are only used sparingly, in applications where there is no option but to use PTFE.

As polymers go, PTFE is not the easiest material to deal with. It behaves contrary to nearly every other plastic and requires special processes to create even the simplest of forms. Whether we look at the extrusion of PTFE tubes or the forming of PTFE films (both rather straightforward when we consider melt-processable polymers), the methods we need to employ for PTFE are a practically standalone and need to be understood and developed from first principles.

Similarly, over moulding a plastic onto a metal part is not a very complex task when we look at injection moulding. In such a process, the metal part is inserted within the injection moulding die and molten polymer is injected and cools around it.

However, there is limited scope to follow this process when trying to over-mould PTFE onto a metal object. The limitations are the following:

1. As PTFE cannot be melted, there is no way to form the polymer around the metal part in any way that would be uniform and consistent using heat alone
    
2. Since PTFE can only be compression moulded, the metal part would need to be kept within a compression moulding die and the dry PTFE powder would need to be packed around it. However, as PTFE is very sensitive to the amount of pressure being applied during compression, it is essential that the metal part does not have too many contours, as this would lead to an irregularity in the compression.
   
   It is possible, in the event of a more complex metal part, that isostatic moulding is used to ensure there is even pressure on the PTFE powder during compression. However, as isostatic moulding requires a lot of die and mould costs, such a process could only be justified if the volumes are significant.
    
3. Even if we do manage to pack the PTFE around the metal uniformly, the final issue remains concerning the heat cycle.
   
   PTFE is sintered (cured) at temperatures of around 375°C over a period of anywhere between 15 and 100 hours – depending on the size of the part. This heat cycle means that within the oven, both the PTFE and the metal are subject to high temperatures. Here the challenge is to match the thermal expansion and contraction of the PTFE and the metal, so that mismatches in the rates of expansion do no cause the PTFE to crack.
    

Overcoming these challenges required a lot of R&D. Different combinations of pressures and heat cycles were employed until we were able to consistency achieve a part that would not crack. The moulded part also needed to be machined, which meant the part needed to be free from any internal irregularities as well.

The best products are usually borne from testing the limits of what can be manufactured. As any company would testify, making a commoditised item has few benefits other than giving a good boost to volumes. The true joy in manufacturing comes from creating something specialised and knowing that the client not only appreciates the efforts involved, but that they would have no reason to take their business elsewhere.

One such product we were asked to develop was a unique and challenging profile in ePTFE (expanded PTFE). We should point out that at the time of taking on this development, our own understanding of ePTFE was still nascent and our knowledge of what our equipment could manage was also still developing. Nonetheless, we took on the challenge, because it seemed like a good one and because, if nothing else, the learning curve it would take us on would undoubtedly leave us more technically sound than when we started.

The challenge

 Put simply, the client had two basic parameters:

• An expanded PTFE (ePTFE) sealing element with a width of 20mm, a thickness of 12.5mm and a step on one corner for fitment (see picture)
• A specific gravity within 0.35

Both these metrics were challenging. For one, making such a large cross section would involve significant load on our equipment. Making ePTFE involves stretching, which needs to be done by gripping the top and bottom of the tape and pulling it. The higher the cross section, the more the load on the pulling mechanism.

Most expanded PTFE (ePTFE) tapes have specific gravities of between 0.55-0.75. Attaining 0.35 would mean stretching at an even higher rate (more stretching means softer, less dense tape) when the cross section is already very difficult to stretch.

Furthermore, the step makes it complicated. We were not sure how the profile would remain after stretching, since the tape tends to get squeezed in order to improve the grip and stretch the tape adequately.
 

The approach

We started by making a simple rectangular profile, without the cutaway. We needed to first assess whether we could even attain 0.35 on a cross section this size, considering the load involved.

At first, it seemed fine. The final dimensions were a bit off, but the tape coming off the machine seemed really soft. We tested a small piece and found it was just under 0.35. However, when we checked the same tape a day later, we found that it felt harder. Another test showed the specific gravity had increased to nearly 0.6!

This led us to our first learning – that because PTFE has memory, it will try and revert to its original form. The stretch rate we were giving was so high that the expanded PTFE was ‘breathing back’ as it was cooling, as it tried to reduce the tension within the material.

We were able to remedy this by using special spooling restraints, that prevented this ‘breathing back’ until the ePTFE had cooled to room temperature.

The second challenge was the step itself. We did not know to what extent the stretching process would deform the step. Making a step exactly as per the drawing would not work, as there was definitely some amount of deformation to be expected. Ultimately, we had to go with trial and error here, which involved a lot of die work, as the extrusion die needed to be modified and/or scrapped to change the final dimension.

The result

We were told by the client that they had approached 4-5 different manufacturers before they found us. All had been defeated by this development.

The resulting product is now a part of a sealing mechanism that has the potential to revolutionise the sealing method used in the client’s industry!