Ever since its discovery in 1938, PTFE has constantly found new uses, becoming an invaluable part of myriad applications. However, despite being around for over 70 years, many misnomers exist around PTFE, with many assuming to be just another polymer and expecting it to behave and be processed in the same way.
As a processor, this often poses problems, with clients unable to understand why PTFE components should be so expensive (even before the price increase), why certain shapes are not possible to make and also why the scrap value has not been factored back into the pricing.
As awesome as PTFE is given its various properties, it is just as difficult to process, handle, machine and even dispose of! I want to look at some of the aspects of PTFE processing and compare them to the myths that I sometimes come across in the industry.
Unlike other thermoplastics, PTFE can only be cold-moulded. That is, you cannot melt PTFE and inject it into a mould to give a desired end-shape. The main reason for this is that PTFE does not flow when heated above its melting point. It attains what is referred to as a “gel state” – where the material goes from being opaque-white to transparent, but retains its shape even in this state. While in gel state, PTFE is soft, but still not completely pliable – making it very difficult to handle.
Given the absence of injection moulding – many conventional and otherwise obvious shapes which other polymers are available in become more complex when applied to PTFE.
All in all, 4 conventional methods exist to mould/ extrude PTFE.
– Compression moulding: PTFE resin (powder) is filled into a die cavity – usually with a simple shape (eg: inner-outer diameter, basic profile, length and width) – and the powder is compressed using a hydraulic press. Pressures would range from 300-400 Kg per square cm. Due to the high bulk density of PTFE, the resin is compressed to a third of the volume it occupies in the die. So if you were looking for a 100mm height, you would need to fill the die to 300mm. Once compressed, the PTFE is left to “dwell” for anywhere between a few hours to a day (depending on the size), before being loaded into a sintering over where the heat finally reaches the melting point of PTFE (about 375 Deg. C). At this point the granules melt and fuse to one another to form the final product.
Given compression moulding can only be done with very basic shapes, the cost of a component/ part gets amplified due to the wastage factor. A simple bowl, for example, would require a block of PTFE to be moulded and the cavity to be scooped out – making for a very expensive affair, especially given that the PTFE raw material is more expensive than most other plastics. Similarly, a film of PTFE requires a special process called “skiving” – where a cylinder of PTFE is rotated perpendicular to a blade, which “peels” a thin layer of film. Again the wastage involved here is high – thereby inflating the cost and final price accordingly.
– Ram and paste extrusion: Usually this is used to make PTFE rods, tubes, pipes and profiles. Extrusion is also used to make thread sealant tapes and expanded PTFE tapes and sheets. Here the resin is blended with an extrusion aid (normally naphtha or Isopar) and pushed through a die at high pressures to give the final shape. Again – the process is cold, with heat only being applied at the final stage to take the PTFE from its “green” (raw) state to its sintered state. Sintering for large tubes can be done in a conventional sintering oven (as described above) while for thinner tubes, a conveyor system can be used, as the tube can be in lengths of up to 500 meters.
Wastage here is again high. A single extrusion requiring a charge of 5-8Kgs would have a 500-800 Gram “end cone” wastage – or about 10%. For long tubes, there is invariably some wastage during sintering as well.
– Isostatic moulding: It is possible to achieve some degree of complexity in shape using this process. The PTFE is filled into a rubber mould which has the desired shape on the inside. The mould is then fitted into a chamber which is then filled with fluid (usually oil). Pressure is applied to the chamber, so that the mould is compressed – thereby compressing the PTFE into its final shape. Isostatic moulding is however not wildly popular because it is an expensive equipment and although it offers savings on material consumption, the payback was not considered fast enough – given that PTFE prices were consistently falling up to mid-2010. It remains to be seen – with the consequent rise in PTFE resin prices – whether this process catches on again, as processors try to control costs.
Even with isostastic moulding, there is a degree of inaccuracy in final dimensions due to the tendency of PTFE to shrink during sintering. It is therefore common practice to keep sufficient stock on the component – which can then be machined to attain the desired final dimensions. Again – wastage is increased here, over and above the degree experienced in injection moulding.
Sintering PTFE is again a difficult task, which many processors initially struggle with. Getting the temperature, and equally importantly the timing of the temperature changes correct is essential to ensure the moulded parts do not crack. A cracked piece is virtually useless – as PTFE cannot be recycled (more on this later)
Timing is crucial – as stated above. After moulding, the PTFE item is kept to dwell for anywhere from a few hours to a full day (depending on the size). The purpose of the dwell time is to ensure any air trapped in the moulded piece can escape, as it would otherwise cause the piece to crack in the oven. Many processors – in an attempt to increase the rate of output – compromise on dwell time, with the effect of having inferior products that will often break during machining.
The timing of the temperatures – or the sinter cycle – is also crucial. While the actual cycle curve is a proprietary technology for each processor, the total cycle time is fairly common across the industry. Small articles (under 100mm in diameter) require only 14 hours in the oven. Slightly larger articles need 24 hours, while very large items need up to 52 hours.
So unlike conventional polymers, PTFE processing is time consuming. A sheet requiring to be skived from a large billet would need about 5 days end-to-end, as moulding would take 3-4 hours, the dwell time would be about 24 hours and sintering would take over 2 days.
While PTFE machining is not very different from other polymers – as far as tool selection, feed rate and RPM is concerned, the material does behave differently during and after machining. A few of the anomalies we have observed are:
1. Tolerance: we often get drawings from the customer specifying tolerances in the range of +/-0.01mm for virgin PTFE. Usually, the designer is someone used to dealing with metal parts (where such tolerances and common and easily attainable) and assumes the same holds good for PTFE. While we have been able to machine glass filled PTFE (which is stiffer than virgin PTFE) to within 0.015mm – virgin PTFE, being much softer, does not allow itself to be machined to tolerances closer than 0.04mm. Again – it may be possible to achieve closer tolerances even on virgin PTFE. But it would require fine tuning of the machining process, and possibly some extra tooling – all of which would increase the machining cost.
2. On smaller pieces, the softness of the material causes it to bend during machining, resulting in ovality and poor finish.To counter this, very small parts often need to be done in more than one operation. This again puts pressure on the tolerance (with CNC machining, a single operation is always preferable in achieving close tolerances) and also increases the cost of the part.
3. Shrinkage: this is one of the toughest attributes to account for during machining. Especially for parts being exported to colder countries, we have received reports of components being out of spec. dimensionally. PTFE is known to change dimension by up to 3% between 0 and 100 Degrees Celsius. So a part with a 20mm outer diameter in India (at about 30 Degrees Celsius) could shrink by 0.2mm (1%) in going to, say, Canada (where 0 Degrees is not uncommon). So with a tolerance of 0.04-0.05mm – the part would definitely fall out of tolerance. It is normally not possible to apply a formula for shrinkage and expect that the part will be dimensionally correct when it reaches the other side. The best bet would be to take a range of 1-1.5% for shrinkage and machine 3-4 sets of samples with varying dimensions and see which one best works for the client.
4. While virgin PTFE is soft and puts very little wear load on the tool, some of the compounded grades are not so easy on the tooling. Our experience has shown that PTFE+Carbon+Graphite, for example, puts nearly as much load on a tool as when machining metal components. So when machining compounded grades – one needs to account for the tool wear out – as it does account for a significant portion of the costing.
As mentioned earlier – clients do sometimes ask whether the scrap value has been factored back in to the costing. It usually takes some time to convince them that there is no scrap value as such when we look at PTFE.
Usually, thermoplastics lend themselves easily to scrap recovery. The scrap is either ground back into granules and can be re-melted and used in injection moulding, or it has some basic scrap value, eg: road builders sometimes add plastic waste scrap into the tar mixture where it melts and adds some strength to the tar.
Since PTFE does not melt, it does not lend itself to either of these processes. In fact, the only known way of recycling PTFE scrap is to convert it into micro-powders – which is an expensive process and done by very few companies globally – and is only applicable to virgin PTFE. So filled grade scrap is worthless, while virgin scrap does sell, but for a fraction of the resin price – making it’s impact on costing negligible.
I hope the above piece has been somewhat useful in outlining some of the nuances of PTFE processing and clearing some of the commonly held beliefs about the same. In case you wish to know more, do contact us via our website: www.polyfluoroltd.com