Got the injector working

I know, it's bad form to reply to one's own post but there it is. The nozzle idea didn't work. I had hoped the fluid would be coming out with enough velocity to deflect and follow a new path. Instead the nozzle just started to dribble. I watched the cylinder piston move though and it moves pretty fast so O-ring seals may be OK. The existing injector orifice plate holes must be drilled at an angle because that's how the jets of fluid come out. The holes are about .008" diameter so I'm thinking about trying to drill them straight with a .012" diameter drill. Maybe if the drill bit is held so that only about .05" is unsupported it will be stiff enough to avoid being deflected too much by the existing holes as they are opened up. Maybe I need to see if I can remove the orifice plate and replace it with one of my own. Perhaps a large hole can be machined in the existing orifice plate and a new one added directly on top that has a hole pattern better suited to applying lube in a straight line. Eric

A good initial assumption is that if the process is inherently linear you need trims for gain and offset. The offset trim corrects for zero errors such as a minimum pulse width to open the valve. You might find that 1 gives no fluid, 2 gives 1 unit, 3 gives 2 units, 4 gives 3 units, etc.

-jsw

Or perhaps you can locate an injector with a tighter spray pattern from a different application.

If its a lubricating fluid,a nd the o-ring is of a material suitable to work with teh fluid it should work fine. Of course you need to size the o-ring groove properly for "just enough" o-ring engagement. I use o-rings for low pressure injection of plastisol (10-30 PSI maybe), and for high pressure retention of air (4500 PSI). Anyway, if you are worried about stiction you can go to a chrome lined hydraulic tube, but I use bare aluminum tube for a lot of applications. Low pressure air cylinders, plastisol injector, etc. Will a piston be controllable and do the job?

I switched my entire shop air distribution over to my refrigeration air drier and I really like it. I have filter seperators before and after with filter seperators on each CNC machine as well, but my main concern is water, not the miniscule amount of other stuff that might get in.

Once you get your 555 circuits figured out (two of them probably) you just need two linear pots to control them.

Is that with the filter back in place? With more viscous liquids, temperature can make a large difference, so keep that in mind, as well.

I'm not sure how small they go and what diaphragms they use (oil proof?) but well water pumps work against a pressurized tank to maintain pressure. The membrane separated the air from the water.

Lincoln had an accumulator in their brake systems on the Continentals way back. Brake hydraulic pressure was run through the steering pump to activate the accumulator, providing assist to the brake master cylinder, giving you a lighter pedal. (Most cars use vacuum assist for power brakes.) IIRC, they were about 6x6", but they might be too pricy for this. I don't know if they could be utilized in a lower-pressure format, or run in reverse (I never took one apart to see the innards), but you could check.

What kind of volume are you going to maintain? What size reservoir?

True. Moisture and oxidation are not your friends.

Why not keep the reservoir enclosed, away from air? I suppose it depends on how finicky your specific cutting fluid is.

Maybe a toothed curve to allow small droplets to spread out the diverted oil as it drips off. What tip style did you say you had?

_That_ you'll have to play with, numbers vs full turns. Run it and measure, charting the output at various temps and viscosities (if those vary). Optimize the pulse width with the 555's RC circuit values, right? That's the fun part, once you have a new device working.

If the cylinder is emptying fast, perhaps you need a much larger reservoir. What kind of flow volume are you working with? ML per sec/min/hr? Ounces?

If you need a certain direction and are having problems with the flow, consider machining a block of brass or AL with those characteristics and insert the injector into that. Perhaps drill and tap it for mister or spray nozzles? The injector would be turned into the flow meter rather than both meter and applicator. If it dribbles at the end of the run, perhaps you could put a valve on the spray tip block or tube.

Thanks for the injector info John. When it is all working I won't be using any type of diluent. The reason I am now is so I can practice without losing any of the actual cutting tool lubricant because it is so expensive. So I diluted some oil with some kerosene to mimic the the viscosity of the real stuff. Eric

The injector you mention above, the Chevy one, IS a throttle body injector. At least that's how they are listed online. I'm a little confused about injectors, I need to do some more research. Eric

It's hard finding the right pattern Clare, there are so many. I really want a flat cone. Like a flat fan. The injector I have on hand does not use a pintle, but I don't know what the internals are like, what kind of valve it actually has. But I'm betting it has a needle valve, as opposed to a disc like the Lucas injectors. It may be possible to machine the orifice disc completely out and replace it with one of my own design. Eric

What make, model and engine is it from, or what is the number on it?

Would it make that much difference? Would it make any difference? How much moisture could be absorbed from the air? There is probably a well known answer, except to me.

It's from an old Acura, 1980 someting vintage. I think. Anyway, here are the numbers: On the side with the + symbol near the connection-35310-23600 On the opposite side-9 260 930 013 The injector is an off white color. The orifice disc has 4 holes in it. There is also a symbol molded in that looks like the stylized H in an oval that Hyundai uses. But I was told it came from an Acura. Eric

A piston will do the job, it only needs to move when the pressure is relieved on one side when the injector is open. The other side of the piston is pressurized with shop air. I already have the 555 thing worked out. I have the pause between pulses and pulse length. So when power is supplied it starts the pulse/pause routine and stops when power is removed. Most of the time the pulse will be fairly short and the pause time long because a tap will be getting lubed and this needs to happen only once per hole. But when milling there will be short pause times. The timer activation and de-activation will be programmed with a M code. Eric

I don't know if direct contact with humid air will cause any problems over time. Maybe not. But since I am building the thing from scratch I thought I would save time and do it now. The reservoir can be any shape so why not make it round and drop a piston in between the oil and the air? With my present system the oil is exposed to air and one type of oil did thicken over time. I don't know if it was because of exposure to moisture. The oil is vegetable based, proprietary, and the outfit selling it wouldn't say whether moisture could be a problem. I am now going to be using a different brand of MQL lube and I just want as many problems as possible worked out ahead of time. I am having fun figuring out how to make the new applicator system. Help from people here has been, well, very helpful. And as a bonus it's all metalworking related. Eric

The part number is for an injector from a Hyundai or Kia. Beta 2 engine and they are rated at 190cc. They apparently actually test at

196cc at 45psi.

You really want a Lucas style disc injector.

see:

I think a 3.8L GM V6 is one source but they are twice the flow rate of what you have.

BMW 325E would be closer to what you have but needs a ballast resistor to run on 12 volts. (Low Impedence)

1.6 liter ford too - but those are not common in America. Some 5.7 Chevys used them too. along with a lot of others.

"Akin to rancidification, oxidative degradation also occurs in other hydrocarbons, e.g. lubricating oils, fuels, and mechanical cutting fluids."

I've been out of the communications contracting business for 27 days now so forgive me for being a little rusty. I forgot to mention there are modestly priced timers available already assembled. The Altronixs 6060 board comes to mind. Yeah they aren't as cheap as the components, but the time to make it up costs me more than the price of the board.

Or just get a cheap arduino and you can even implement "peak and hold"

I already have the timer made Bob, but thanks anyway. Eric