Extra Information

I worked in the plastic machining industry for over 10 years before I chose to setup my own machining business. At first I was going to do what I knew best, machine plastic, but there was a problem, supply. I am originally from the UK and there is no issue with sourcing these materials, after-all there are the largest suppliers in Europe. Canada and America only supply some of the basic shapes which means ordering from Europe. Once the shipping costs are factored in it makes it almost impossible to compete in this industry. I decided to stick with general machining while I get myself in a healthy financial situation and work on a solution to this problem in my spare time.

I want to concentrate on the sealing industry as I have a lot of experience in seal machining and design, I also have designed a lot of tooling specifically made to making these parts accurately and fast. I used to machine hundreds of pounds of polyurethane a day back in the UK with seals up to 36 inches in diameter, I was lucky to machine 10lbs a year in Canada. Polyurethane is used for hydraulic seals due to its durability but with that they also require Acetal or Nylon components, again these size shapes being impossible to find.

Another material I machined hundreds of pounds of was nitrile and FKM. Rubbers have an extremely high viscosity when it comes to molding, but what makes injection molding impossible is that they vulcanise under temperature, in around 2 to 15 minutes. So compression molding is often utilised to make these billets, however a pressure of 15ksi is required to mold them. Sizeable billets require extremely expensive molding machines and molds. There is another method as explained later that I want to utilise.

PTFE is chosen for its very low coefficient of friction while being soft. It is often used for wear bands and sliding surfaces, however this is not its best use, actually this is often a bad use due to its softness. Seals with spring energised lips is by far its best application. PTFE has some very favorable high heat performance properties which can be further improved by filling with carbon. Other fillers such as bronze, glass and moly can reduce friction and enhance wear properties. PTFE is very similar to rubber in that it flows poorly and sets under temperature, so again has to be compression molded. I'm hoping I have another solution to this, but it requires a little bit of experimentation beforehand. 

PEEK was 95% of the material I machined here in Canada. PEEK is one of the highest performing plastics out there but also comes with a very hefty price tag. Since the price was so high it actually negated the cost of the shipping somewhat making it fairly profitable. PEEK is not without its issues however, again suppliers are very few between, especially on anything over 3 inches in diameter. The other issue is that suppliers can't seem to perfect making good billets. Injection molding often leads to air pockets and a lot of internal stress in the final parts, this making dimensional accuracy difficult to achieve but more importantly a potential structurally flawed part. Compression molding often produces better quality billets with no pockets but even more internal stress. Lastly there is a centrifugal casting method which is used to produce larger diameters, unfortunately there are only a couple of suppliers in Europe able to do this. All of these billets are barrel shaped which makes manufacturing very difficult, you don't want to bounce a $1000 billet around your machine.

I will be aiming to become the largest manufacturer of large shaped engineering grade plastics. More specifically materials that are used in the sealing industry in the form of rods and tubes. There should not be a limit in terms of diameter but initially I will go with what I can successfully machine on my current machinery, which is approximately 18 inches in diameter. I also wish to offer small billets since someone might only want to make a single part. (Ok I cheated, the image above is AI).

Like any business capital is often the most important part when starting or expanding. I literally started from nothing when I created Machined Plastics and have slowly increased my machinery capabilities. I still have a lot of legwork to do on my current business in terms of financially securing the unit I am currently renting. Thankfully this process can be started off small as the biggest cost would be having the machine manufactured, which is something I can do myself. Even with the capital at this initial stage I would still start small, from what I can find there are no machines for sale in the methods I want to produce this material. I call it centrifugal casting but it differs quite substantially from the methods used currently.

Once the small prototype has been built its a simple case of running controlled experiments. There are a number of variables, some of the basics ones are material density, melting point and centrifugal force. There will come to a point where the data starts to correlate and predications can be made, and confirmed. Once a particular material or size of mold has been proven then another variable is added. I have all the range of materials to go through, but data of melting points, densities and viscosities are readily available.

I believe once data has been collected on three types of materials and three different sizes of mold then I am in very good stead. To get to this point it may require me to go through about 1500lbs of plastic material, hence starting with cheaper plastics. Testing the sample billets also requires destruction to prove if there are flaws, which I hope to achieve a 99% success rate. Even the materials I machined myself from the premium manufacturers were not devoid of errors, so I'm hoping to be more consistent with my process.

This is the point where I can try to attract investors here in Canada or I can see how business goes by myself first. It gets exponentially more expensive to build the molds the larger a billet is, thankfully the molds themselves are fairly basic and not too expensive to make. 

I used to design a lot of stuff in my spare time but did not have the machinery to make it, but now I have the machinery I don't have a lot of spare time, or rather spare money. I had a few days off over Christmas and started on a design for a sequential transmission. I have a BMW M70 V12 engine I would like to put in a project some day, so I would like to machine every part and make all the electronics myself. I don't want to be the norm, so making my own transmission is a very important step.

I decided that I should make a completely mechanical sequential this time. I started again with the ratios and made sure they had the right spacing. 3.86   3.00    2.39    1.93   1.57   1.29   1.00    0.78    0.62   0.50   0.41   0.33 . I feel like these give the best of all driving types and doesn't make the gear shifting too ridiculous. Take note that gear 4 is traditionally 1.00 and gear 5 is the overdrive, this transmission has five overdrive gears. It is similar design in that there is a primary set of high / low gears coupled to six sets of gears. There are few variations on dogs, either both on the engagement ring and gear, or in my design just on the engagement ring. My gears each have a cutout in them to accept the dog, this does make it stronger at the expensive of shifting reliability. Better slam those gears!

With this new design I can save some space with the cutouts in the gears and I also saved some space with less bearings. If a helical gear is used then the gear is quieter and stronger at the expense of higher side forces. Having side forces means that taper roller bearings have to be used. Going with straight cut gears removes the side load and a regular deep groove ball bearing can be used. Straight cut gears are very slightly more efficient at the expense of being weaker for their given size and considerably noisier. Overall I managed to make the gearbox to half the length of the original design.

I now own a wire EDM machine which I can use to cut the straight gears and the splines, it may take a while but it will be worth it. For scale the diametric pitch of the gears is 5 inches, with each gear being 1.5 inches wide. The largest gear is a 60 tooth with a diameter of 7.75 inches. Each bearing that presses into the gears is of 85mm / 3.45" in diameter. I probably won't be running big power through this setup, maybe 500HP, so it should be durable enough.

One part of the design I was stuck on was the shift forks for the dog engagement ring. Most of the time these use a basic fork that gets lubricated by the splash of oil, just the aluminum acting as bearing material. The friction is only ever present for a very brief period when changing gears, but my design is a little bit different. I did not want this part to wear out too quickly as potentially the output shaft could hit 15,000 rpm or more. Needle roller or even ball bearings take up a lot of space and are fairly limited at these higher speeds due to the diameter. The ring itself is about 4.5 inches diameter in my design.

I did a lot of thinking again with the layout and the design. Should I make it all mechanical or air-assisted for faster and more reliable shifting. I also have one ring between two gears, but what if it was to engage just a single gear. I set to working on a completely new revision of the gearbox while keeping the same gear designs and ratios. I came up with a spring loaded pneumatic cylinder for each of the gear sets. If there is a lack of pressure then it will be in the neutral state and only one pressure valve is required for each gear set. I have dropped the all mechanical aspect in favour of making it simpler.

I will include some hall sensors on each cylinder so that it cannot engage another gear until the actuator has returned.

I'd say the design is very close to being completed, my only issue is that I won't be able to completely 3D print a plastic prototype. The cylinder and piston will likely need to be made from metal, or even I machine them from plastic.

Once I am happy with the 3D printed prototype it would then need the lubrication system worked out, and that I want to go a little crazy on, I want this thing to last. I was originally thinking that the middle shaft will be coupled to a large flowing hydraulic pump, that can then be used to lubricate every single bearing and gear. The problem is that this shaft can reach 18,000rpm with an engine speed of 6000rpm, so I feel that rules out a mechanical pump without further gearing. There is also the issue that when in neutral the middle shaft won't be spinning even if either the engine is or the output shaft is. I also have an idea to couple an electric motor to the middle shaft as part of a hybrid / generator system. So while I don't particularly like the idea of it, for now it might utilise an external electric oil pump. Splash lubrication is enough for pretty much every manual gearbox out there, my only worry is it getting to the bearings on the middle shaft.