Hybrid Hammer

[daccel]

I doubt anyone remembers this project from..wow almost two years ago now, but it was my unsuccessful attempt to build a single impact pneumatic chisel. Well, not a complete failure I got it to work a few times, but it was too heavy, cumbersome, unreliable and difficult to service. Have I sold you yet?

Enter christmas holidays 2009... Back for a week to visit the folks and bringing all my tools and spud parts (strictly as weight for traction I tell my wallet) so I can putter away in the shop. Time for version 2.

Have decided to try and make it hybrid... the only way to get enough power in a compact enough form without a higher pressure source than a regular compressor.

In a very basic sense, it consists of a tube with a chamber on one side of the piston and a captive barrel on the other. There is a second piston between the chisel and first piston which allows for pressurization to return the first piston. That much I'm confident I can build, complete with o-rings. Probably using 2" aluminum with a 1" bore drilled for the cylinder. This will allow for any slight error in boring and plenty of wall to tap inlets, outlets etc.

Now it gets difficult . The layout I have pictured should work for fueling and returning the piston. A series of spool valves are connected to the chisel so that when the tool is lifted up, fuel fills the meter pipe, and the piston is pushed to the top of the tube. When the tool is set down, air flows through the fuel meter pipe, pushing fuel then air into the chamber. Spark, bang, repeat.

Now where I'm stuck is on how to suspend the piston at the top of the tube until ignition, and at the same time ensure the air in front of the piston is evacuated fast enough to not form an air spring.

Two possible solutions I've thought about are pictured below. One would be a pin that springs into the piston as it moves up. As the chamber pressurizes, the pin is blown out of the piston and the piston is released. Only I just realized that same pressure will put significant side force on the pin which could jam it.

Another way would be to have a QDV at the bottom of the barrel, with a switch built in. Momentarily after the barrel begins to dump, the chamber would ignite. HGDT doesn't show much performance loss due to ignition at time of release (entered as low burst pressure in model) because of the heavy 'projectile'.

Any ideas on that would be much appreciated. And any other aspect of it of course, the aluminum rod is pricey so I thought I'd run it by you fellas in case I'm out to lunch!

Edit: had a moment immediately after posting - connect the barrel to the chamber with an air spring loaded spool valve and set the spring pressure to just above the preignition pressure, so that the barrel dumps before the piston has a chance to go anywhere! Did a quick sketch, I think that just may do the trick.

[Willdebeers]

I would personally prefer to just use HPA, as it would be so much more convienient. But I do have to give you kudos for drawing up a very detailed drawing. Bravo!

[jackssmirkingrevenge]

Have decided to try and make it hybrid... the only way to get enough power in a compact enough form without a higher pressure source than a regular compressor.

Surely since available area (and therefore force) increases exponentially with piston radius then surely you can get power from 150 psi to do the job without having an exceedingly large piston?

It could be as simple as having the piston and chisel with spring return in a chamber fed by a pop-off piloted QEV, varying pop-off pressure would adjust power and varying flow rate adjusts your er... chisels per minute

[POLAND_SPUD]

I agree with JSR that you should build a pneumatic version instead... but the design he suggested has a couple of problems... first of all, if you attach the cylinder to the exhaust port on the QEV air couldn't escape from the cylinder

you have to add vent holes in the cylinder body so that it could vent excess air when it extends fully and either make the piston/air cylinder leak on purpose so that it could vent air as it comes back or find another way to vent air


I guess that what you used in the past already had large enough piston but the flow was too low

with such a large piston/air cylinder you really need at least a 3/4" QEV valve

you could pilot it with a 1/4" manually actuated 3 way valve (or smaller but then you would have to add another smaller QEV (1/8" or 1/4") to pilot the large one)

[psycix]

I'm thinking you are making this too hard. Its a problem I also have:
I design something, but a few things are not right, so I design something else to make sure it does, and that makes me end up with a way to complex design full of workarounds, small pistons, pushing rods and hammer valves to solve problems.

The idea is nice, but I think you should simplify (or go pneumatic).
Pull the first piston back with a spring, and leave the 2nd piston out.
Give the first piston a smaller porting (like a "stepped" piston) so that the spring can hold the preignition pressure (small area)
The high pressure after ignition opens the piston, and when the piston moves back, it gains extra surface area for the pressure to push on, ensuring full thrust.

[daccel]

I think onboard hpa would be consumed too fast and don't really like the idea of 3kpsi lines in a rock yard with heavy machinery , but point taken on simplicity.

You don't have to increase piston diameter significantly to get the same force (for example going from 1"x1" chamber 1"x9" barrel to 2"x9" chamber 2"x9" barrel is similar) but it begins to approach sizes which are hard for me to source materials for and to machine. On the last one I sleeved pvc in steel pipe for the barrel, because there can't be exposed pvc for strength, but it didn't stay in place well and was a less than ideal solution. And making it all out of steel at those sizes or slightly larger it was too heavy. Though obviously you want some weight to reduce recoil. So not ruling out pneumatic, but I don't see it working with my budget/sources/tools, but I will give it more thought, could be missing something.

Popoff QEV would work, but piloted would be more appropriate. I'm not trying to replicate a jackhammer, but need a single strike on demand, for which I haven't been able to find a commercial solution. Return spring sounds good though, the loss in force would probably be worth the gain in simplicity.

Same with the vent holes below piston and leak above, two less things to go wrong. That was part of the problem with v1, (beyond my lesser spud skills at that point ) while it would work in theory and occasionally in practice, there were too many components that could malfunction.

Me? Complex solutions? Never . I like the stepped piston idea. I kept shying away from spring return because I was worried about energy loss and not being able to hold preignition pressure, but that addresses the latter. The relatively high piston weight should minimize the loss from restricted flow.

Assuming I go hybrid, any thoughts on fueling? I was thinking regulate to 50psi propane with a meter pipe so that it can operate down to 0 degrees, but maybe it would be simpler in terms of the spool valves (is that the correct term btw?) that control the fuel and air on/off to build a low pressure regulator. Oh and I just realized I have an accurate scale which measures up to 300g which I could use to calibrate the regulator without buying an expensive pressure gauge that is accurate at low pressures, sweet!

edit-spelling

[kenbo0422]

On the last one I sleeved pvc in steel pipe for the barrel, because there can't be exposed pvc for strength, but it didn't stay in place well and was a less than ideal solution.

Did you try setting the PVC inside the steel sleeve with set screws of some kind? With the steel on the outside, boring into the pvc a little shouldn't pose a lot of pressure problems. Also, lock tite, believe it or not, would help.

[daccel]

The pvc was ~3/16 smaller than the steel, so I used some foam tape to center, then some screws to hold it. I suppose if I just used more screws I could get it to stay put. Not a lot of wall to screw into though, and would worry about deforming the pvc if it was just set screws pressing against?

Here is what I have in mind for fueling. Not to scale and the regulators aren't shown. Trying to keep it simple for the operator so they just have to place the tool and pull the trigger. I think if the valves are large enough with a high enough flow rate and a small chamber, the chamber should fully pressurize at each step.

[ramses]

would a nailgun style safety be acceptable? This could be used to control fueling easily with two linearly actuated valves, one 3 way. Maybe 3 for a pneumatic auto-return. A 5 way in place of one 2 and one 3 way would simplify things

Valve one would have one port tied to the regged propane input, the common port tied to the meter pipe, and the final port tied to the chamber via a high pressure check valve.

Valve two would control air.

Operating mechanism:

Code: Select all

On pressing the safety, valve one would close off the chamber, and fill the meter pipe. Valve two would fill the chamber with air.
The optional "valve three" would open the atmospheric port, and close off the low pressure source.

BOOM!!!! :D

On releasing the safety (which should move slowly, so the "recoil" doesn't cause misfuelings) Valve One would dump the meter into the chamber, and cut off air. Possibly the 3rd valve could close the atmospheric vent, and dump air at a low pressure to re-set the piston, which would be locked by some kid of friction device.

rinse and repeat

This is all based off "hammer.jpg" in the first post. some kind of "cocking" procedure would be needed. The A BV porting the chamber to atmosphere could accomplish this (Open BV, depress safety, close BV, release safety, read ) Or you could use a microcontroller, and pulse solenoid valves, like previously mentioned and commonly implemented in paintball guns to control the bolt/spool valve.

EDIT: I just read your post, and you have the exact same idea. (skipped to the pictures ) Back to your actual problem, you should most likely make some kind of piston valve/QEV to pneumatically hold the force. You wouldn't even have to pilot it, if you made the pilot area large enough. It would be a good idea if the "hammer piston" moved far enough "down" to vent the area behind it to atmosphere. The last image you posted would put a lot of stress on the spool valves, as you would effectively be actuating them with a sledgehammer, unless there was a disconnect, like in a QDV.

Also, LINKEY

Just make the spark "semi auto" and you're pretty much set. Although it is standard combustion.

BTW, I wouldn't trust foam between PVC and steel to transfer the force well. the brunt of the force would be supported on the tips of the screws, so unless your barrel looks like a pincushion, bring on the epoxy!

[POLAND_SPUD]

the spool valves (is that the correct term btw?)

technically they are Directional Control Valves (because they direct flow from one port to the other) but they are also refereed to as 3 (but also 4, 5 etc.) way valves.

In Europe it's common to use a slightly different designation, for example '3/2 valve', '5/3 valve' etc. the first number is the number of ports while the second one tells you how many position the valve has

Spool valves are probably the most common type of DCVs... it's basically a DCV which has a small spool that slides inside the valve body... I often used this term to refer to DCVs because:
1. I didn't know much about them in the past (sorry I am learning as I go)
2. it's convenient




this hammer is basically just an air cylinder... so a single acting version could be controlled with a 3 way valve while double acting version via 4 or 5 way valve


the only problem is the size of it... with 2" piston you really need a lot of flow to achieve high velocity

most 1/8" Directional Control Valves have a flow of about 250 - 600 L/minute;
1/4" - most around 1200 L/minute
1/2" - the one I found achieves 2400 L/minute

and even 2400 L/minute is not enough to open the cylinder fast

I've modelled your 'hammer' in fluidsim to see what flow rates you need

I used the following data:
pressure - 120 psi,
'hammer' moddeled as a rod less air cylinder - 50mm piston diam. & 50cm stroke

the results:
even 2400 L/minute was too slow (it took 0.2 second for it to fully extend and that's with friction and mass set to that of commercial air cylinders). the results are probably better than what you could achieve


QEVs could offer higher performance - I could only find data on the type of QEV common in Europe (link to data sheet)
as you can see there even 1/2" QEV offers almost twice as high flow as a 1/2" DCV but you hsould really try to get 3/4" version

I guess that it might not be a bad idea to buy a 3/4" QEV and put it on the hammer you already have

[ramses]

Poland: The DCVs are used for metering only. I doubt that he will ever need any more than 10L/sec, especially considering the semi-auto stipulation. Therefore, 1/8" valves should be adequate. Anything else (the vent) should be a QEV or a custom spool valve (maybe 1.5" max diameter of the spool.)

The actual hammering force comes from essentially a (hopefully valveless) hybrid. I am convinced that a valve would make things much easier, and flow restrictions will be less important because of the high temperature (therefore SOS) of the gas. The weight of the "projectile" would also reduce the effect of flow restrictions.

[POLAND_SPUD]

yes you are right that a 1/8" DCV might be enough for metering but I was referring to my previous post and JSR's suggestions to keep this project fully pneumatic

I suspect that his pneumatic hammer didn't work well because of insufficient flow and because it was too complicated and cumbersome to use

He could get higher speed and therefore power if he mod what he already has. He only has to use a larger piston valve or a QEV/QEVs

edit
It's not that the setup I proposed uses a lot of air - I've just modelled it again: 50mmX50cm air cylinder with a 3/4" QEV works pretty well with just a 1.5L tank at 120 psi. But surely it would work better with 2 X 3/4" QEVs or a 2 - 2.5 " piston valve

I know that flow rates that I mentioned in my previous post can make you think that it will use a lot of air but in fact it's just like a pneumatic cannon - you need high flow only for a fraction of second

[daccel]

Thanks for the input guys.

In your defense my posts can be wordy, I would probably skip to the pictures myself .

Well you could lose the upper set of stops on the chisel in the last picture I posted to prevent the chisel's downward movement from ripping the valves open.

$2k for a fence post driver? Maybe I should consider marketing this!

Hmm DCV it is then, though I think I've seen spool valve referred to more often around here.

Maybe it would be worth trying to find a pneumatic cylinder and modify it to take a larger QEV in place of the small ports. Then you'd have your piston too. Or even a small hydraulic cylinder for the hybrid. Will have to look around. Found this on craigslist, but seems a bit overkill!

I don't want to reuse the old cylinder and piston, because the piston was poorly cast from bondo, the orings aren't floating, and I would rather not sleeve pvc if possible.

The part that was least reliable on the old one was piston return and retention at top of stroke. Originally had a 1.5" piston valve which I couldn't get to function reliably, so I switched it out for a 1" sprinkler valve just to test the whole system.

Here is a more complete drawing of a hybrid version. Tried switching the fueling actuation to be based off the combustion. But now thinking about it, it is probably simpler to have it based off chisel position, just design it in such a way that the valves are protected. Oh well at least I'll have paint skills after this .

Edit-when you say you are convinced a valve would make it easier, you mean a standard piston hybrid with a second piston (hammer) in the barrel? I think a return spring if it didn't sap too much power in combo with vent holes would be easier.

Edit2-here is what the rock looks like fyi, this would be used for splitting into 1-3" slabs for the most part. There are weak layers that allow it to split like that, with some encouragement, but too much and it will just break into small pieces which aren't as valuable, which is why they are stuck doing some of it by hand.

[CS]

Everything looks good, stepped piston face (good idea I must say), but like you mention I see problems with driving the fueling cycle directly from a piston as pictured.

Only thing is the piston reseting, perhaps it won't fully return close?

Anyways better then I can explain it:

[ramses]

Long story short, I was looking into controlling meter position based on pressure spike a while ago, and I theorized that it couldn't be made to work. The reason is really hard to explain, but I don't think it will work.

When I said to use a valve, yes, I meant make a piston hybrid. Don't use a pilot valve, instead, hook up the regged down air to the (large) pilot area through a high pressure check valve. I would say just use a spring, but IIRC, in Fnord's first piston hybrid, the spring he used behind his valve (for essentially the same purpose) got pancaked. Also, to restrain ,say, 40psi across a surface of around 3 square inches would require 120 lbs of force. Once the huge, expensive spring is compressed, it will exert a lot of force on the piston, your most recent design looks to have some anti-pancaking built in, but I would rather not buy, and support a huge spring like that.

For fueling, a 3 way is not required for air. A two-way dumping the regged input will do just fine. That is to say that you should "manometrically meter" the air or O2. If you use O2, you will most likely need a meter pipe, unless you feel like a 15x hybrid minimum.

Also, I think metering based on a nailgun type safety would be a good idea. Something like this could be implemented. You could still use the roller valves, but come up with a ground down rod to actuate them by way of linear motion. It would be much easier than attempting to seal roller ports inside the chamber of the gun, as shown in pimpmann's diagram. No offense...

It would be a very good idea to make the system modular, to deal with different types of chisels, and to replace broken ones.