Manual injector design help

[Cogsy]

Over the Summer I need to build a top secret apparatus for an experiment I'm assisting with at university and I've run into a design issue. I need to be able to manually inject something with the consistency of toothpaste (OK it is toothpaste) through a 2mm (0.080") hole, into an expanding matrix of 2mm square channels. I figure it will take quite a lot of pressure to properly fill the space.

So I was thinking of a simple nicely lapped aluminium piston in an ali bore, with maybe an o-ring to seal it, with a handle as a lever driving the piston. What do you guys think? Will it provide the required pressure or do I need a better arrangement?

Of course I also need to build the 2mm square channelled matrix but I'll worry about that once I get the injector worked out.

 

[stevehuckss396]

Dont know what the budget is but i would use a motor and ball screw instead of a lever. The thrust would be huge. Attach the piston to the screw and rotate the nut with a motor and belt.

 

[RonGinger]

If it is really toothpaste remember that stuff is abrasive- some guys have used it as a very fine lapping compound. It will likely wear out any piston cylinder rather fast.

 

[canadianhorsepower]

I would simply try a grease gun
you can generate pressure up to 6 000 lbs with a good one.
we use that to remove seized king pin

http://www.machinerylubrication.com/Read/28803/grease-gun-pressure

 

[Swifty]

A seal like in a grease gun or hand tyre pump may be better than a plain O ring, the more pressure applied, the more it seals the bore.

Paul.

 

[petertha]

You mentioned the project is coming from a college. I'm guessing the problem would benefit by even a rudimentary flow simulation rather than guessing at things. Probably software like Solidworks can model this, or get you in the ballpark. Most colleges & universities have discounted full featured license seats available to students & staff. I'm guessing the plastic injection molding industry evaluates applications like this routinely before they even begin a mold.

In your case, sounds there are no temperature cooling effects like plastic injection, but just the ability to specify viscosity & flow passage geometry might help yield understanding injection pressures, maybe fill-up issues to ensure branches aren't incompletely filled etc. I can almost visualize when you have one feed going into a network of other channels, things can get funky. Sounds interesting, good luck.

 

[Buchanan]

Why don,t you take a standard medical syringe with out the needle and test some standard toothpaste in it to see what force is required to get it to flow.
You could easily add a brass tube with a nominal area the same as your channel and the same length. this would give you some basic facts to work with. You may need to drill out the orifice of the plastic syringe body but it would e a quick and easy way to find out what you need to know. You could hold the syringe pointing upwards and push it down on a scale to read the force.

 

[Cogsy]

Thanks for the input guys. I'll look further into the grease gun idea, sounds like it might be the way to go (no real budget so motor and ball screw can't happen).

I will be attempting to simulate the flow in a program called MATLAB but it's a little more complex than I described so the physical modelling should help us create a better mathematical model in MATLAB. If my supervisor allows me to, I'll post the results (as long as it works).

 

[kvom]

Drill a .08" hole into the cap of a toothpaste tube and see how easy it is to squeeze out.

 

[Cogsy]

Drill a .08" hole into the cap of a toothpaste tube and see how easy it is to squeeze out.

Now why didn't I think of that! Thanks.