Designing pneumatic injection molding machine

Hello guys.
Martin from Slovakia

Im currently designing pneumatic injection machine with molding force from the  pneumatic cylinder 4,7 kN. Diameter of cylinder, where plastic will be melted is 40 mm, height is 170 mm.Do you guys think its enough force to inject plastic into to mold with keeping good quality of the final plastic product ?

To control the pneumatic cylinder im using in my design 5/2 valve. To control this, i can use PLC or Arduino. Its a bit more complicated, but could be quite useful, and save some time, if used in a right way and settings.

The construction is designed from welded closed stell square profiles 30x30x3 mm.The maximum volume of injected plastic is 224 cm^3 , which is quite satisfying for me.

The advantage of this machine should be quicker production, and in right setting also better output of the surface, thanks to quite big pressure (around 470 kilograms, with operating max pressure that is allowed allowed) if im calculating correctly.

Disadvantage of this, is a bit higher price of making this machine, which is approximately 1000 EUR, and also a bit more complicated operation.

The model was made by using SolidWorks software.


Im aware, that some mistakes could occur even in this model, since its my first time to work with pneumatic systems.

On the picture you can see theoretical model with some section view, and also with injection simulation of smartphone cover.

What are guys your thoughts about my design. ? Thank you for any ideas or tips.Cheers


I am not sure that is enough force for that cylinder diameter. My manual, arbor press actuated machine has a similar diameter. The arbor press is 1 ton. With the extended handle, I’m pretty sure I’m using a good portion of that capability pushing HDPE through an 8mm nozzle.

In terms of saving time, the vast majority of my time is spent heating/compacting the plastic. Once ready, the actuation is almost instantaneous in comparison.

so I’ve put for fun put a 75cm diameter steel wheel/pulley on the on my CNC vise handle socket and moved it with the fingers: insane strong, stronger than me closing the vise for milling Aluminium ! I think even a Nema23 could turn this with some plastic printed gears/pulley with ease. About making the wheel, you could extrude a beam straight through a ring roller, close it up and add the teethes with a DIY mini shaper or just extrude the pulley profile already… The inner supports of the wheel or gear can be done in a couple of million of ways. sorry, it wasn’t pneumatic but I think it get’s the job done nicely, no idea how much pressure in tons this makes …

impressive! I was wondering it’s easier to use a 100E CNC vise for the clamping; with a little fiddling one may automate this with a stepper motor, a little expensive but still more effective and cheaper trying to get same results out of a pneumatic system though.




Below are some pictures of the clamping unit from my 2nd machine, you can see a video of it in operation here


I prefer the over-centre toggle method to a screw, it’s what nearly all industrial machines use, it’s faster and applies all the force at the end of the stroke where you need it, though a linear screw clamp is simpler to construct.


This clamp applies around 5 tonnes I think, though I don’t have a good way to measure this accurately. It’s powered by a cordless drill motor with a large gear reduction, all spur and planetary gears for maximum efficiency. There was quite a lot a machining involved in making it, I had to make all of the gears plus some more you can’t even see. The two arms are supported in pairs of heavy duty bearings inside the plates.


Oh nice, its coming together here! Indeed, the machine is a little bit bigger than planned in the beginning-right now 1,8cm high+ a table to rest on ~2,5m

Which clamping system do you use right now? And do you know under which pressure the mold-clamping fails?

I had some problems with simple screw-based clamping before, but it was always more or less easy to solve based on the closing torque of the screws.

Do you want to share a picture as well maybe?

Hey Martin,
I’am working on a similar concept since December, i looked out for many oldschool machines and youtube concepts and went through several iterations in CAD design. I will create a new development post for feedback for the next iteration, once the build is done.


Hyrums machine on youtube is really nice, another nice concept, but harder to find, due russian:

Facts shortly:

– 28kn by a 200mm Pneumatic Piston with 300mm stroke

– 3 exchangable barrels with inside diameters of 20mm, 26mm and 34mm to vary pressure (311-900bar) and volume (270-85cm³)

– sliding barrel to accomodate for different nozzle heights

– semi automatic machine, clamping will be developed in the next step

– currently made from 30mm Aluminium Plates, but ready for 20mm Steel lasercut

– control via ESP32+Display (Nextion?) or ESP32(via Node-Red)+Smartphone (for easy PID tuning, statistics and later automated control)

Still a lot to improve on my concept, especially manufacturing wise. I would really like to switch to threaded rods too like Jannis’s concecpt, like pprog pruposed. The Plunger connection in my design looks a bit sketch too.

I’am curious to hear how your plan was to use an Arduino for contrlling the machine, did you planned for a display as well?

Maybe you can remove your top-frame (beyond the flange mounting holes of the cylinder) and use two solid bars to mount the cylinder (unless you plan for a housing). Or you could screw the cylinder to a steel plate which is under those two bars.

Did you made a force simulation for the frame? On my last concepts the fixture was always a little bit the weak point of the structure. The strength of the profiles decrease drastically as soon as you drill holes through them!

Also make sure that your balance between injection pressure and volume is selected well. If you plan to make phone cases i can easily see that you want a higher pressure and a smaller volume.

So how i understood this relation it is that the injection chamber diameter is correlated to the injection pressure

( Force(N) / (r(mm)^2*PI) )*10 = Pressure (bar)

That the nozzle will reduce the final pressure is sure, not exactly sure how this behaves for plastic as well, due the main factor might be the melt flow index (MFI) for polymeres and its correlation to the speed of molten plastic.

I will do some further research in the next weeks here in v4 to explore the behaviour of plastic in molds. Maybe we get some insigths from that 🙂

@s2019 I was forecasting pressure in the nozzle diameter via simple pressure calculation → P=F/A , and i was expecting the increase of the pressure, with a decrease diameter, as this simple equation is saying.
However i checked your link, and also i checked Bernoulli equations brefly, and its quite opposite 😀 , there is dicrease of the pressure, with dicrease of the pipe diameter, when something flows inside.

Thank you for the input ! Atleast i know now, that i need to recalculate it 🙂

Sorry, I don’t see why the pressure in the nozzle would be higher than the pressure provided by the 40mm piston.

The chart in on this page illustrates it qualitatively

Jannis posted today a nice way to keep things together via ACME rods; the video suggests the same; so you can end up with 4 such rods and 2 plates (or bars) just. a rework of normal 20T presses could give a perfect framework btw. too.
let’s see tomorrow what can be done about the clamping links 🙂

@pporg  you are welocome, hope it will be able to open it.
Yes I saw this video as many similar to that. i didnt catch btw book recommendations.

And yes , im aware that it is a first design shot, still very raw. 🙂

thanks a lot, did you check the video entirely ? it’s all there nicely done there and said, incl. an awesome book recommendations. your frame for instance can be lots of simpler and cheaper 🙂

@s2019 , yes your calculation is correct. Its pressure approx. 3.7 Mpa.

95 MPa, should be created pressure in smallest diameter of the nozzle, where it connects the mold runner.

Here is 3D model link , →  Free CAD Designs, Files & 3D Models | The GrabCAD Community Library
that has lot of mistakes i know. Its just theoretical model for now, tested just in simulation. So i would very happy appreciate some improvements. @pporg
Hope it will work. I used function “Pack and go” so everything should be there 🙂

Because i think, if you design this right, it can save to of time and maybe some energy ( that you dont need to press it by human power) to make it semi-automatic.
Thats the whole idea


s2019 is right, the pressure you get in the barrel is the pressure you will get at the nozzle, the change in diameter makes no difference. An easy way to calculate it from your air pressure driving the pneumatic cylinder is:

P= line pressure x  (dia. cylinder)² / (dia. barrel)²


So if you have a 100mm cylinder at 8 bar (0.8 Mpa) you will get approx. 50 bar (5Mpa) Which is acceptable and probably a little higher than the PP lever operated machine. Less than around 30 bar and you will start to have problems and will need to heat the mould to get good results.


I have built 2 pneumatic semi-automatic injection machines (now building a 3rd) and I try and aim for 1000-2000psi  (65-130 bar). I’ve found the biggest problem is the clamping system which has to be able to cope with the pressure in the mould, which multiplied over it’s area can result in much higher forces.



200mm is a huge cylinder! I look forward to seeing that!

Isn’t 4700N acting on a 40mm piston about 3.7 MPa?

@s2019 I done injection simulation in Solid Works Plastics with all parameters from designed injeciton molding machine, and after all, the machine was able to create pressure in the nozzle more than 95 MPa, which is quite good, and also simulated injection of cover was done without complications.

But im sure in order to have more comfortable operation and also variety of use, it would be cool to use stronger pneumatical cylinder for this design.

I will try to upload the model somehow.

@pporg in my design, the nozzle and the mold is connected with screw connection, similar as it was originally designed by Dave for injection machine.

And yes, i was thinking a lot how to design the hopper in appropriate position for smooth inseting shredder plastic into the machine.
And yes, i think the production price for this kind of design can be squeezed into 1000 euros.


traditional, semi-automatic pneumatic injection systems as seen on youtube (my fav) seem to do things a little differently … i guess you already watched out for it … the second difficult stage is the clamping system; in both cases, there’s lots of detail work to to figure. the more simple has the clark 25 desktop but the solution is in the yt video.

that’ts great news btw. somebody from the PP world picks this up 🙂 might it be possible to share this to a 3d-online viewer ? that way others could assist easier …

if you need help with coding or hardware, I am all yours 🙂

The hopper could be a little more direct and less disturbing, a cylinder coming from the side (at 45 degree ?); in a way it can be heated and pre-compressed (with auger or pneumatic plunger) and thus, automated as well.  like a carousel
tool changer, just loaded with cartridges …

about the prices, i’ve seen lots of pneumatic stuff for sale at second hand sales, can be cheap after all; we buy this in kg so that isn’t really an excuse 🙂

@mod, there’s lots of machine work to play with online, what’s best to do now ?