Square section buffer and ram?

[Liftoff]

Lets for a moment forget certification and testing of pneumatic parts... would it technically make sense to build a square ram and buffer tank?
The reason behind the idea is to save space, and build a more compact machine. The buffer and ram can be bolted together as a single piece and will contribute to the overall rigidity of the robot.

The setup would run on Low pressure, say 10bar and material I'd use is extruded aluminium alloy - box or rectangular section for both ram and buffer.

I know box section is not ideal as a pressure vessel and there will be a tendency for the sides to bulge, but is it worth exploring the idea?

[typhoon_driver]

Off the bat I can think that machining would be an issue. I have no idea if the box section would be strong enough. The issue you have there is the flat sides. A round cylinder is inherently stronger.

By all means give it a go as long as you are safe about it

[maddox10]

Compared to a round ram/buffer it's idiotic heavy if you want to make it safe.

Also, getting the piston to seal needs a miracle.

My idea is to use 2 oval rams stacked and a buffertank made of several small diameter pipes squeezed between 2 end plates with a lot of small tierods.

[Liftoff]

Compared to a round ram/buffer it's idiotic heavy if you want to make it safe.
Also, getting the piston to seal needs a miracle

True but how much heavier? The only way to find out is by trial and error so I'll get a selection of different gauge box section extrusions and check how much they expand when pressurized, the local scuba service centre should be able to measure that, they have hydro-testing equipment.

Getting the piston to seal should not be a big problem, and even if it were to leak slightly the loss would be negligible. The seal I had in mind was a piece of 5mm nylon or similar material cut square the exact size of the bore, sandwiched between two plates to make the piston. It might leak from the corners, but it wont have time to leak much given the high flow from the buffer/QEV to the ram.

Most compact buffer / cylinder system would be to have a concentric buffer around the piston

Not sure I got this right Gary...you mean the cylinder has a double wall and the space between the walls is the buffer?
Might be an idea...actually a very good idea... and I think you solved my problem

Mario can you please elaborate on the problems of such a setup...

[typhoon_driver]

Most compact buffer / cylinder system would be to have a concentric buffer around the piston

[Max]

Most compact buffer / cylinder system would be to have a concentric buffer around the piston

This is what I'm planning, along with a massive QEV built into the base of the ram on full pressure! should send bots flying if I can get it in the weight limit.

[maddox10]

Tried that, wasn't a succes. Too powerfull and fast. Have the photo's to prove it.

It's to much work for the results. Even for me who can make all the parts at home.

But if you want to build such a setup, please do. I like to be proven wrong.

[Mouldy]

Tried that, wasn't a succes. Too powerfull and fast. Have the photo's to prove it.

It's to much work for the results. Even for me who can make all the parts at home.

But if you want to build such a setup, please do. I like to be proven wrong.

After you sold that setup to grant cooper, it got totaly re-machined, weight taken out, re-bored, new plunger and all tolerances made properly, it now works reliably and its fitted into satans mutt.

The buffer ram system works great.

[typhoon_driver]

Yes that's exactly what I meant. I believe the robochallenge guys did something similar for their satans mutt robot for the gadget show although I don't know the details.

With such a setup, there would be little point going to full pressure as Mario says, the system is just too powerful and would likely destroy itself in the process unless it was over engineered to the nth degree (titanium this and that). Of course, the ideal "valve" mechanism would be a plate on the bottom of the ram that has a number of "petals" machined out. A corresponding set of petals on a rotating plate with some form of actuation to expose or close up the holes. Sealing such a system could be interesting. I'm just thinking out loud now though.

The other issue you always have with featherweights is the low weight limit relative to the power of the rams.

[Liftoff]

It was never my intention to go FP with such a system, max pressure I want to work with is 10bar. As for the valve, I think I'll use what I already have (which works fine) a QEV bolted to the bottom cap of the ram.The QEV could also be enclosed inside the buffer. At 1/2" it should be enough flow to feed the ram.
To be honest I'm not after an increase in power, but rather to have a more compact setup, which should (in theory) leave me with a lighter robot as I can build it smaller.