epoxy pump

[darryl]

Replaced my nearly used up epoxy today. Again I'm thinking about some kind of positive displacement pump to use with it. I want something more elegant than 'squeeze the bottles, gauge how big the puddles are'. Currently I'm keeping the bottles upside down in a holder, so there's no wait for epoxy to flow to the tips, etc. It works- I just want to progress to a better dispenser solution.

Some years ago I came up with a dispenser for epoxy that happened to be quite thin stuff. It was gravity feed- worked, but no guarantee of equal amounts coming from each feed line. I still had to weigh each component to ensure I got an equal amount of both. With the common 5 minute epoxy it would be too slow anyway. What I want is to be able to pump a handle and for each actuation a fixed amount of both components would be ejected from their separate nozzles. Each nozzle would have a spring-loaded valve that would open under the pressure, then close to prevent seepage. I do not want to mess with a cap each time.

The twin parallel tube dispenser is a step in the right direction, but doesn't go far enough to satisfy me. I'm willing to build something that will do the job.

So far the spring-loaded nozzle idea looks good, and would only need a method of introducing the epoxy to it, then plunging a precise quantity out the end. A piston in a tube is the basic way to do this, and this is where the fun starts. If you have to pull the piston back to introduce fresh epoxy before pumping it out, then you risk getting air past the seal and interfering with the action. Without a fool-proof sealing method, this type of pump could be problematic. The big question here is what kind of seal would work well like this- keeping in mind that this is a highly viscous compound being pumped. If it was oil it would likely be a no-brainer.

Another way to do it would be to enhance the twin-tube method with smaller diameter, but probably longer tubes, and you'd pump bit by bit one way until they are used up- then you would refill them. Maybe this is a better way to go, mainly requiring that you have a more precise mechanical control over the two piston actuators, plus add a reasonable method of refilling them. Some kind of ratcheting mechanism could be used to 'crank out' a fixed quantity of both components for every actuation. Seems reasonable.

Another idea I had was to make the piston itself a one-way valve, then have the supply component above it. During retraction the valve would open and air-free epoxy would fill the nozzle area below the piston. Epoxy would feed in through the top and keep the whole mechanism filled at all times. The requirement then would be that the pistons seal well enough that there's little seepage past it as you're ejecting the current 'charge'. If both components leaked past the pistons at the same rate, you'd still have an equal mix coming out the nozzles and you'd be ok. The actuating rods would be immersed over most of their length, and long enough to allow the feed containers to be full without the rods pulling epoxy up and through the top bushing. A slight gap at the bushing would be fine as some air would have to enter the system to allow epoxy to exit. You might use a seal there and provide a one-way air intake valve- the deal here being that there would be no 'breathing' and possible epoxy smell coming into the room.

Lastly for now, the other concern is the choice of materials to make this out of. It would all have to be non-reactive with both epoxy components. Steel cylinder material, actuator rods, and maybe even the pistons could all come from nitrogen charged hatch struts- if the steel/epoxy combination would be ok.

Brainstorming online here with this idea. Add (or subtract) what you may-

[boslab]

I tried using one of these washing up liquid dispenser bottles with a silicone rubber valve type thing on the top for gorrila glue recently, seems to be working well so far, epoxy however is sticky viscous stuff, especialy when cold, at least araldite is, i used a twin cartridge dispenser once but it worked out very expensive, i would be interested to see how you go about it
Mark

[lakeside53]

I used piston type pumps in the past - and have abandoned them. They are cheap to buy from epoxy suppliers, but unless you use a lot of epoxy it isn't worth it - they tend to clog, are hard to dispense with many epoxies, and are messy (some portion is exposed to the air and tends to thicken over time).

For small quantities I use weight. In the case of JB weld, the hardener and resin both have the same specific gravity so a digital postage scale works for perfect 1:1 ratios. For other ratios or SG, you might have to do simple calc.

For larger volumes I use pretty much anything lying around. For a 2:1 mix, I fill the container twice for one part and once for the other. Pill bottles work great.

[TGTool]

I'd wonder about using a peristaltic pump. Twin flexible tubes would meter the same amount with the same amount of rotation. That doesn't solve the drooling problem, but you might make the rotation notchy and times so that the rollers stop right close to the end of the tube. Of course if you have check valves following that wouldn't be necessary.

[elf]

Print one of these on your 3D printer, mount it over a scale, control it with an Arduino or Pic.

[garyhlucas]

A twin tube peristaltic pump will work very well if you make it reversible. Pump out what you need, reverse to push it all back in the bottle. I happen to sell industrial ones and many of the them don't even have rollers. They just have sliding shoes, lobes on a shaft and they lube everything up with silicone grease. A rotor with three lobes would work great as a manual pump. The big pumps adjust the occlusion (squeeze) of the tube with shims under the lobes. Once it is set no need to adjust again.

You could also look for stackable peristaltic pump heads on Ebay and just make a manual drive handle you hand crank.

[darryl]

I've also considered something like a peristaltic pump- what I had in mind was a flexible chamber that would be alternately squeezed and then released. One way valves would direct the flow of fluid. You would have valves but no seals and pistons. During the release phase, the chamber would have to return to shape via its own power of elasticity. Having that 'softness' as it were, it would likely be more subject to detrimental effect by the fluid it's handling.

You could use a diaphram, and then of course you'd have to make sure it could 'inhale' and expel a fixed volume of fluid for each activation. Could be done I suppose.

In either of these methods, I've thought to use a hydraulic fluid on one side of the device- that would be controlled with a piston and cylinder. The difference would be that the piston would be working in the hydraulic fluid and not the fluid being pumped. Still though it seems like a less precise method- maybe I can think about this more.

A long time ago I considered how I could use surgical tubing as part of a vacuum pump- again the idea was to remove the need to have seals and pistons involved in the actual pumping. The idea was to contain and pressurize a short length of surgical tubing inside a cylinder. Ideally, the shape inside the cylinder would be exactly the shape the tubing would take when inflated. Then when you let the pressure out, the tubing would collapse and draw a vacuum behind it. As a pressure pump it would be inflated to force the volume between the tubing and the cylinder to evacuate through a one-way valve. Releasing the pressure would cause the tubing to shrink, drawing in a fresh charge of whatever is being pumped.

My experiments along these lines used typical 1/4 inch surgical tubing and lengths of about six inches. I found it would inflate up to about one inch in diameter before reaching a point where it 'got tight'. From this I determined that I could have it expand to about 3/4 inch (inside some 3/4 inch ID tubing perhaps) and it would remain undamaged over many cycles and a reasonable length of lifetime. My last experiment with the tubing was to continue to inflate it as hard as I could until something let go- and it did. It blew very loudly and disappeared. All I had left was two ends and a ringing in my ears.

In a very conservative constraint of operating parameters though, this could make a good pump. Likely the tubing would be more affected by the fluid when it's inflated, and in any event it would have to be tested.

[firbikrhd1]

Not a very exotic answer, but 50 cc syringes are available as are 100 cc units. A couple of those, one with each part of the epoxy and you can accurately meter out what you need regardless of the ratio. Clean 'em up with alcohol and use 'em over.

[GadgetBuilder]

You can purchase pumps and epoxy from West:

http://www.westsystem.com/ss/calibrated-pumps-make-accurate-measuring-easy/

The pumps aren't expensive but the ratio of epoxy to hardener may not match the epoxy you're using. You could get two epoxy pumps and work out the ratio for your epoxy.

The West epoxy works well, commonly available at boat yards where it is used for fiberglass repairs.

John

[lwalker]

One of the standard ways to do this is with replaceable syringes in controlled slides. Each slide moves a measured distance to dispense the desired amount of fluid. This allows you to create infinitely variable mix volumes. Imagine a holder for the epoxy syringe that has a leadscrew and a stepper motor to drive the syringe piston the distance needed.

The machines I work on (I write the control software) will normally use a single piston or syringe pump with solenoid valves to allow aspirating fluid from multiple locations. Since I'm an electrical/software guy, this is easier for me to visualize than a purely mechanical method.

We also use peristaltic pumps, but they aren't as precise.

Now the question: after you have a precise fluid dispense, how do you mix it?

[J Tiers]

If you are not doing this on a production basis, why go to the trouble?

Most types are available in dual piston dispensers, and there is no real requirement to use the mixer tubes if you are not using really fast cure stuff, and are careful about mixing by hand. Even the box stores have dual piston dispensers for some varieties.

You wanna talk about waste.... all the stuff between the peristaltic pump and the nozzle that sticks to the walls and won't pump back.... For non-production-volumes, the whole approach seems like an unbelievable mess and hassle.

Suit yourself, though.

[darryl]

Waste in this case would be what's left in the pump and lines at the time I'd be throwing the whole thing out. Otherwise as I refill the 'tanks', the old gets used up before the new comes in. I would have to make sure I always put the supply in the proper tanks

I don't have a production environment, but there are often times that I'd mix up a dozen or more small batches in an evening- probably do this on average about once every week or so. Mixing is easy as I just poop it out onto a non-absorbent surface and stir it around with a toothpick- usually the same tool that I'm applying it with. That's not likely to change. I just want easier metering of small volumes and not have to mess with caps. Must be related to OCD

[Paul Alciatore]

It seems to me that the biggest problem with epoxy dispensers is one or both of the components tend to harden somewhat when exposed to air, even small amounts of air. This tends to say that you really want a new, fresh dispenser after a period of time, perhaps a year or so. In any case, it would be a very good idea to keep any air out of the reservoirs.

And, I have never had any real problems with matching the amounts of equal part epoxies. The modern epoxies seem to be tolerant to small mistakes in the proportions. I wonder if you are seeking a solution to a problem that does not really exist.

[superUnknown]

http://www.adafruit.com/products/1150

$25 peristaltic pump. Use them for accurate reagent dosing, they'll pump up to 14 PSI when run on 12VDC. Considering a Cole Parmer peristaltic pump is $1200, these are amazing little geegaws.

[boslab]

How about a couple of small pistol type grease guns, the grease delivery tube is usually rigid but a shorter flex cut in two would suffice those things can squeze viscous stuff?
Mark