We've just launched our map. Add yourself by clicking here!


Plastic Plate Press, semi-DIY/semi-professional

Mark Bachrach markbertbach

Plastic Plate Press, semi-DIY/semi-professional

25/01/2018 at 12:23

Hi, I’m Mark, An Industrial Design Engineering student.
Currently, I’m working on my graduation project, the design of an Open-source plastic plate press for bottom-up recycling in low resource areas.

I’m planning on finishing v1 of my design before the end om May, this year, but a lot has to be done still.

in a few weeks I’ll start the real design work, and around the end of march I’ll start building a prototype.

But first I need to find out what people expect from such a machine, this is where your help comes in. If you are someone that would like to recycle plastic waste through use of a plate press, please help me and tell me what I need to know. To make it easy I made an online survey: https://mark171.typeform.com/to/yrgyP5

I’ll keep you posted on the results, developments and creations in future posts.

25/01/2018 at 12:23: Up til now I started collecting and small-scale experimentation with a panini iron.

07/03/2018 at 11:11: I’ve done a range of experiments, Gathered survey results from around the world, set a goal and now I’m doing actual design work. scroll further down this Topic to see what I’ve shared so far and feel free to comment, all help is very welcome.

03/12/2018 at 18:42:

since today, all of my documentation is publically available here

I did this project for and together with the MMID Foundation so make sure to attribute to them and me if you do any publications, as it says in the Creative Commons Attribution 4.0

This nicely rounds off this project and since I am now working on the sheet-press at V4, still sponsored by the MMID Foundation, I’ve started a new Topic to keep you posted on all developments.

96 replies
18 subscribers
17 saved
sort on most likes
12/06/2018 at 13:58

Yes, very impressive !
What plastic do you recycle?
What is the material of your molds?Do you use a separator?
We did some tests with PET – the most difficult to recycle, but the most interesting for its UV transparency – with aluminum molds, we had to use teflon!

12/06/2018 at 16:21

awesome work !!!!!!!!
love the heating down system with the little ventilator ^_^
what thickness is your output product? veryyy nice coulouuuurs !!

13/06/2018 at 11:12

@richardtim & @imuh & @andyn

The result has a nominal thickness of 12mm, but there is variance between 11.8 and 12.8mm

The mould material is sheet metal 1.5mm cold rolled steel, and the frame with the cutoff edge is made from 10x10mm hot rolled steel with a 13 mm strip of cold rolled steel sheet  2mm thick welded to it

Grinding with 300gk sanding disk makes the sheet metal very shiny and smooth. Finishing it with PTFE spray or Vaseline spray helps with the release but for the lid ‘ilI need to improve this, I guess the release agent just drips off during melting.

So for the top side, I’ll try PTFE sheets
Plates are 770 x 1080 x ~12 mm = 10.4 kg HDPE or 9.7 kg PPPower consumption data WIll follow, but I have to figure that out, which I’ll try when I’m happy with the mould. Hopefully next week.

The ventilator makes a huge difference!! I think the cooling time is cut in half or even better. The rectangular steel tubes have a lot of surface area and work like a heat sink!

19/06/2018 at 18:33

Hi I made another video of the machine building process
Would be nice If some of you could take a look at it and give me some feedback, just imagine that you also have the technical drawings. 🙂


20/06/2018 at 20:26

Hey @markbertbach !
Love how you timelapsed the whole process, also good choice of music, and the output you show is so nice ! Congratz !
I have to say you are sexy at 8:40 (are you naked under that pretty dress? :p )
If i can make some constructive advice, you should raise the sound of the voice when you speak or lower the music at that moment because it is a bit difficult to understand
At around 7:30ish it is hard to read white text on white background 🙁
Very nice work !

21/06/2018 at 09:11

Thx Nick!

It was a very hot day so I thought, why not go full commando 😛

I’ll definitely improve the things you mentioned and I’m also adding shots in between to point out what parts are going to be made and what is important to know about the function. And I’ll add a part about the Hydraulics and Pneumatics and also the Mould Lid.

Comming up in the next weeks!
Thx for taking the time to give me some feedback!

21/06/2018 at 16:26

Superb work !
Congratulations for all the work and for the video !
For the design, did you consider the use of the hydraulic tools from Trad4u,  @olce had mentioned ?
I like hacking tools. Doing everything from A to Z is interesting, but we have to be really good at welding as you are !

25/06/2018 at 11:48


Hacking tools is of course really nice, but in this case, I think it would make the machine more complicated and more expensive. If the welding is a problem, then with some small adaptations, the machine can also be bolted together, that would even be better for critical load safety. But I limited the number of bolts because of their expensiveness and sourcing difficulties in low-resource areas. Welding gear is found almost anywhere.

And thx for the compliments on my welding, but this was actually only my second big welding project, and at the start of the project I couldn’t weld that well. So I guess anyone can learn by doing!! Just make sure to save the critical welds for last, then you have a lot of practice time 😀

25/06/2018 at 18:11

Yes, @markbertbach, you’re absolutely right. It was explained to me that the MIG / MAG welding was quite easy to learn … For the hack of TRad4u it is only practical if the format is suitable, which was not your case!

18/07/2018 at 17:33

Awesome work @markbertbach !

I was wondering a couple of things:
– Do you have an estimate total cost for the finished build?
– Do you have any estimation on the cost of each sheet based on the energy required to make it e.g. in shredding, melting and then pressing?

19/07/2018 at 11:22


At the moment I’m finalising my graduation documentation including such an estimation. I’ll post it in the first week of August. If I forget, please remind me 😅

19/07/2018 at 20:53

Thanks for sharing all this!

Version E is exactly what I was planning to do – i think a large vat would allow for faster production and likely more energy efficient. Could even use various heat sources – rocket stove to incinerate other garbage, for instance. Would take some effort to fine tune the temperature though – may a liquid heat exchanger or something.

I’ll definitely take some inspiration for the press from you though!

19/07/2018 at 21:30


I have to warn you about this option!
I think it will take forever to melt such I big lump of plastic because of the low heat conductivity of plastics. Probably the outer layers will burn begore the inside will melt.
Ofcourse you could ad a stirring device, but this will mostlikely be a troublesome device because of the viscosity of the plastic and the rotor will get very messy.

But it would be great to explore if all of this is really true 😉

20/07/2018 at 05:03

Thanks for the feedback!

I think I’m going to go ahead and tinker with my idea for a while though – I really want to make this cost efficient for the developing world, and I assume electricity is a major cost in this process.

I’m going to try to do some sort of rocket stove heat exchanger – use the stove to efficiently incinerate (organic) garbage (they burn it all anyway) and heat some sort of oil, which will then slowly (and hopefully somewhat stably and accurately) heat up a very large vat of plastic (with some sort of stirring device inside), which I can then use to fill one mould at a time. I figure once its all melted, it will stay melted almost indefinitely and can just add enough for a new sheet for each one poured. Ideally it would result in a fairly quick turnover between moulds and a reasonably high output of sheets that we can then do a lot of cool stuff with.

Efficient/clean cookstoves are part of my project anyway, so might as well tinker with both at once!

I’ll be sure to post progress here in a new thread!

21/07/2018 at 00:35

Hey there, I have a slew of questions/thoughts now that I’ve reviewed the posts and videos a few times. Hope you have time for them!

I noticed you previously used a book press. I was thinking of doing something like this with a single jack. How did that go? Why did you switch to the 4 hydraulic jacks? Was the pressure insufficient?

What sort of pressure are each of those jacks capable of? Do you have any rationale for what pressure is needed or did you just choose those at random? Do you think the rig you built would be able to withstand more force?

I bring this all up because any literature I can find on compression moulding seems to say a minimum of 200psi is needed, which would be over 200 tons of pressure for a sheet that size, which is obviously absurdly out of reach for us. However, most literature on compression moulding is for thermosets, as thermoplastics seem to not be done this way, so maybe that makes a difference. I’m just trying to get an idea for what sort of pressure I should try to build for.

You said somewhere that the thickness of the sheets is variable, which suggests you just allow the pressure to do its thing rather than have some sort of stopper in place to make an exact height. This is probably the correct way to do it, as limiting the height would limit the pressure on the plastic and presumably result in a worse product. However, you also said there was some thin flashing, which is typical of a compression process, but also means that the pressure on the plastic is literally leaking out.

I suppose a real industrial process would use a specific measurement of plastic and the mould design would allow for flashing while also keeping the correct pressure – perhaps just a trial and error process for the likes of us?

Have you tested the sheets in any building processes yet? Do you think they could benefit from additional pressure? Could you get away with much less?

Or am I overthinking all of this and its a forgiving/flexible process? Whatever the case I’m just going to tinker with some scrap metal and see what happens. But I’d like to have at least some rationale for what I’m doing.

21/07/2018 at 10:58

Hi @nickchomey

Thx for your interest and eagerness! and sorry for my absence and brevity.
Next Friday I have to hand in my Thesis so I’m a bit preoccupied 😉

But I could post my entire thesis when it is finished, I think, I’ll have to check to be sure If I’m allowed, but I think I am since I wrote it.

But for now, I’ll copy paste some sections  to answer your questions:

Force requirement:
Leonard S***g suggested this pressing-force requirement, to be at least 64 Ton or 640 kN for a surface area of 1.32m2, which translates to a pressure of about 0.5 MPa or 5 Bar (see appendix XX). For a plate of 1,22 x 1,22, this would mean that a pressing force of 720 kN or 72 Ton is needed, which is quite a lot.

The Plastic plate press system of Heat-MX uses a total of 64 Ton or 640 kN of pressing force, this only result in a pressure of  0,2 MPa or 2 bar. For a plate of 1,22 x 1,22, this would mean that a pressing force of 320 kN or 32 Ton is needed, which is already less than half of what Leonard suggested.

When calculating the pressure that is applied by the press setup used in the prototype tests (book press), the lowest pressure requirement is found. Only 0,175 MPa or 1,75 bar is applied according to the calculation found in Appendix XX. Although this is the lowest pressure requirement suggestion, the calculation is believed to more likely result in a higher pressure than the reality than a lower one. Most importantly, is the fact that this setup does result in plastic plates that satisfy the requirements.

Like most decisions in this design project, the easiest executable and most affordable solutions are tried first and only when tests provide indications that the design can benefit from added complexity, a new solution is sought after. Thus a pressure requirement of a minimum of 0,18 MPa is put in the POR. This results in a force actuator output requirement of at least 268 kN or 26,8 Ton.

Hydraulic jacks:

This choice was made because of the relative ease of instalment which allows the parts to be connected to it to stay as simple as possible. Also, the system seclusion of hydraulic cylinders results in a low need for maintenance and cleaning and the completeness of the cylinders ensure that no additional parts are needed besides the cylinders, opposed to for example a spindle, which needs custom fitting solutions for both the end nut as well as the guidance nut. All of these arguments combined also result in a lower cost price, which is also one of the primary decision drivers.

These benefit can however only be realised if the right cylinder can be found. The cylinder needs to fit the force requirements, the price range and needs to be source-able all around the world.

The minimum pressure requirement has been calculated in the previous chapter to be 1,8 bar which results in a force requirements of at least 268 kN for plates of 1.22 m by 1.22 m and a force requirements of 158 kN for plates of 1.10 m by 0.8 m, which is the size of the mould that is going to be prototyped.

Since the design goal is to be able to produce plates of 1220 x 1220 x 12 mm, the higher requirements are chosen to fulfil. This will allow for the prototype design to be scaled up without a need for other cylinders.

The price limit is difficult to decide upon, but as a first suggestion a fourth of the total allowable cost price is granted since the force actuator is 1 of the four main parts; $ 1200 / 4 = $ 300.

With these requirements in place, the search was set out. International websites were used to look for hydraulic cylinders, and soon it was found out that the most affordable cylinders are bottle jacks, which are often used in auto garages to lift up cars or press bearings in fittings. Also, longer bottle jacks were found that are used in engine hoists. In the analysis, it was also found that car parts or garage tools are an excellent source of materials since cars and garage tools are global common goods, and thus are good for the source-ability.

Press application

In the press, a stroke of 10-20 cm would be enough to provide access for placing the mould and pressing it shut since the current mould design will only be 45mm tall and when filled with plastic, will most likely not exceed a height of 10 cm. However, because of the cleaning access requirement as discussed in the previous chapter, it should be possible to move the press beds apart for at least 30cm. Other means than jack extension length can also realise this but, it would be very convenient if the cylinders could realise this feature.

The same holds for air pump option, which is demanded by the design requirements, but it is a very convenient option that adds to the usability of the product. Next to that, when it is chosen to use multiple jacks, the activation can be centralised by linking the pneumatics.

When it comes to force-output, single small jacks can be found that satisfy the needs, 30-ton jacks or 32-ton jacks would suffice and are available within the price range. However long ram jacks are not found for the required force output.

When considering these three arguments, logic would demand a single small jack of 30 or 32 ton, since only they satisfy all the requirements, although not the wishes. However, the other options only satisfy the wishes and not the force requirement or at least, not by themselves. When multiple jacks are used, also the long ram jacks can satisfy all the requirements. For example, when four 8-ton jacks are used, 32 tonnes of force can be realised.

The choice between a single powerful jack or multiple weaker jacks thus needs to be made. This consideration of multiple jacks may seem un-logical since four weaker jacks are more costly than a single, stronger, jack. However, there is a good reason to look into this option. The force distribution from 1 central point to a large surface area has different structural requirements than the force distribution from 2 or 4 points to a large surface area. Figure XX depicts two simplified situations in free-body diagrams (FBD) with corresponding shear force diagrams (SFD) and bending moment diagrams (BMD). The distributed force indicated with Wp represents the pressure that is put on the mould, FA represents the force generated by the single bottle jack of 32 tonnes, and  FB & FC each represent two bottle jacks of 8 ton totalling to 16 ton each. This doubling is done because of the simplified 2D representation in which the 3rd dimension is compressed in a single layer.

In the diagrams can be seen that the maximum shear force and maximum bending moment can be significantly decreased by choosing multiple load points instead of a single centralised load. Keeping the maximum bending force as low as possible is essential in this design since bending deformations in the press-beds will result in thickness deviations in the plastic plates. These deviations may not exceed the defined tolerance requirement of +/- 1 mm, and thus the bending deformations may not exceed this limit. This can only be realised by integrating sufficient stiffness into the press beds and thus lower bending moments allow for less stiffness which allows for smaller beam profiles and less material.  All of these consequences result in less material use, less complexity and lower costs which all are very much wanted.


Four 8-ton long ram jacks with air pumps are chosen as force actuators as long as costs allow it. This choice is made based on the following arguments:

Multiple load points are beneficial to the design as opposed to a centralised load point. Extended ram jacks have a long stroke (more than 30 cm), which allows for easy cleaning access positioning. Air pumps allow the activation of the jacks to be centralised and by it, the jacks can be operated by a single person. Four jacks, each with a maximum lifting force of 8 tonnes, total to 32 tonnes of pressing force which satisfies the required 26,8 Tonnes 8-ton long ram jacks are commonly used in engine hoists, making them  source-able almost anywhere on the globe Most of the available 8-ton long ram jacks have two single axle fixtures, each on one end of the device, making them very easy to install.

8-ton long ram jacks can be bought locally in the Netherlands for € 70,- per piece, totalling to € 280,- for four units, which is just below budget. Ordering form china, similar units can be bought for below $ 50,- per piece (alibabba.com).

Figure XX shows a picture of a typical 8-ton long ram jack with an air pump, more information about such a hydraulic jack is found in user and safety manuals such as the one by harbour freight:


Fleshing, cut-off-edge & stopper:

In the mould frame iteration tests, subsequently, three frame edges were tested. The tests started with the most straightforward frame edge, a square section all around like depicted in figure XX-A. This test pointed out that the heating and pressing method functioned quite well, but the frame edge did not do an excellent job of cutting out the plate shape and defining the chosen plate thickness. The pressure build-up was insufficient, and thus the resulting plate ended up too thick, and the fleshing was not cut off, as can be seen in figure XX-M.

With the previous test in mind, a frame edge was fabricated that had a very small cut-off-edge, 2mm wide, while still providing enough strength and stiffness by having a fatter base, as can be seen in figure XX-B. This small cut-off-edge was theorised to build up more pressure since the contact area was decreased significantly. This theory proved to be true since the plate result that was outputted, was cut-out perfectly and the plate thickness turned out as planned, see figure XX-N. However, during fabrication of the frame edge, it was noted that it was much work with semi-advanced machinery to produce the specific section and therefore it would not be suitable for production in low-resource settings.

The third iteration was a simplification of the small cut-off-edge. The edge was integrated into the mould-tub, by only folding up a small edge that directly functioned as cut-off-edge, see figure XX-C and XX-F. When filling the mould with plastic flakes, a first problem arose; The granulate towers high above the small edge and is not properly contained and thereby falls out of the mould very quickly, see figure XX-I. When the plate was pressed, however, the mould seemed to do its job quite well. Only when the plate was finished cooling down, and it was taken out, the second problem showed itself; because the previously existing enclosing mould walls were now missing during the melting process, direct heat radiation burned the plastic, as can be seen in figure XX-L. Next to these two problems that disqualify this design solution, the edge itself did do a great job of cutting out the plate and defining the plate thickness as can be seen in figure XX-O.

Concluding, the small cut-off-edge concept has proven to do a great job of cutting out the plate and defining its thickness. However, the producibility should be improved to make it suitable for production in low-resource settings. This improvement will be realised in the embodiment design and will be tested in the large-scale prototype.

I hope you can make something of it without all the images and context, but for now this is all I can Provide 😅 Next week I can be more elaborate and we could even call/skype

23/07/2018 at 02:28

This is amazing – I don’t know how many lifetimes it would have taken me to figure all this out. Thanks so much.

I’ve got a bunch of questions again and I’m sure you’re quite busy, so no worries if you don’t respond right away

What was the 5 Bar suggestion based on?

Does it change for different types of plastic?

Surely the thickness of plastic matters too – 1mm would presumably need much less than 50mm, regardless of surface area. Is there some sort of scientific guideline/formula for it?

You settled on 1.8 Bar based on success at that pressure with the book press.

What criteria did you use to judge it to be a success? Just visual inspection? Lack of bubbles? Proper shape? Actual stress testing the material properties?

Did you try less pressure to see if it would work? More pressure to see if it would be better?

I ask all this because I have been browsing all sorts of plastic manufacturing textbooks recently and they seem to say, in accordance with intuition, that a visual inspection is a very poor and misleading way to judge output quality – differences are at the molecular level and show up in stress testing.

Ive got other questions and thoughts on the jack, frame design (steel thickness requirements), moulds etc… But will leave that for later.

Good luck with finishing the thesis! We definitely need to chat when you’ve got time – there might be a job in it for you scaling this up in the developing world as well as helping us tinker with other stuff to help improve lives in a cost effective, creative and empowering way

03/08/2018 at 11:43

HI Guys,

I just wanted to showcase the latest results, after a redesign of the mould I have been able to produce Beautifull plates from PP(Blue) & HDPE(Black).

also watch the new and improved video!

07/08/2018 at 15:20

Really nice work man!

It looks like you changed up your mould quite a bit, how come? From aluminium -> mild steel?

07/08/2018 at 16:33

wow very nice mark….respect.  what is the next plan.

07/08/2018 at 16:34

I would be happy if I had this machine in my workshop

11/08/2018 at 10:01


thx for the compliments

I changed the mould to steel because of the source-ability of the material, in some low resource settings aluminium is hard(er) to come by or very expensive.

Also, the mould needed additional strength and welding was the preferred joining technique.

next to that the scratch resistance of steel is better than aluminium which is nice when cleaning mould. And steel thus holds its polish for a longer time.

I’ve added some pages from my design documentation in which the mould design is discussed, evaluated and improved

11/08/2018 at 10:06


Thx for the compliments!

AT the end of this month I have my graduation presentation, around which I’ll also discuss how the design is going to be shared and what will happen to the machine that I built. Afterwards, I’ll maybe join the precious plastic Army in Eindhoven to work on V4 and improve or make a new press design.

And I’m also looking around for a job 🙂

14/08/2018 at 16:36

Thanks for all the info @markbertbach

Currently building a smaller version of your mould for some 300x400x5mm sheets. I’ve made a couple adjustments based on your large aluminium mould and now tweaking it for mild steel with some DIY cut-off edges. Will post in this thread how it all works out 🙂

23/09/2018 at 11:44

Awesome documentation!

10/10/2018 at 13:58

I’ve tried to make a much smaller version of your sheet mould..

It was quite a finicky job jigging and welding the thin parts together but I managed..

Unfortunately haven’t had the chance to test the mould yet..

17/10/2018 at 17:43

@timslab looking good!
i’m looking forward to seeing your plate results!

05/11/2018 at 21:26

Awesome job!  Have you tried using PET to make a panel yet? How many panels are you able to product per day?

06/11/2018 at 15:17

I’m currently working on tests and improvements for the press at Precious Plastic V4 In Eindhoven.

In the ideal scenario in which you have 3 moulds, you’d be able to make 4 plates in 8 hours

PET is on the test list but haven’t tried it yet

I’ll start updating here again as soon as I’ve found noteworthy insights and results

12/11/2018 at 12:43

Wow, truly amazing work! Feeling really inspired by this 🙂

It would be very interesting to read your thesis, did you ever post it? Also did you post any technical drawings of the machine?

Thanks for this, and good luck with the v4!

Viewing 30 replies - 61 through 90 (of 96 total)

You must be logged in to reply to this topic.

Support our projects on Patreon so we can keep developing 💪