Need some hydraulic input MIGHT BE GETTING SOMEWHERE

[rodekyll]

In the course of building the trike, I've been re-inventing the Convert.  I tossed the gearbox but kept the front hydraulic drive section.  Where the clutch basket used to attach to the torque converter output shaft I've attached a forward/reverse gearbox with no gear reduction -- all of that will be done via the torque converter.  So there is no clutch or gears -- just a hydraulic coupling between the engine and differential.

So far this part is working.  I can start the engine and make the wheels go this way or that way.  But the fluid isn't doing its job.  Symptoms are weird.

The hydraulic fluid loop starts at the bottom of a radiator that has top and bottom tanks and about 4x the cooling surface of the oem cooler.  The radiator is located a bit farther back than the oem unit -- about at the end of the top horizontal frame tubes.  So there is more length of hose than oem (~18").  The radiator cap has a 15# popoff like a regular car.

From the radiator the fluid takes the traditional route through the pump and fluid drive assembly and returns to the tank.  Again, the hoses are a bit longer than oem.  I am using Gates transmission hose in all locations.

Symptoms are:

I am not getting a good t/c lockup -- massive slip.
Tiny bubbles in the sauce
If the cap is off the radiator I idle at about 5-6# atf pressure.  Pressure does not want to come up much until the engine revvs, and then it's reluctant and low.  The wheel motion is proportionately slow and weak, and I can stop the wheels with no rpm drop.  At idle the wheels don't spin (rear axle in the air).  I can see circulation in the tank.

If the cap is on the radiator the pressure builds to about 17# idle and the wheels spin weakly as I rev -- but no power -- no lockup.  If I rev the engine I get better spin, some lockup, and what appears to be rapid heating of the fluid.

If I pop the cap (engine running) I burp out some fluid.

If I pop the cap (engine stopped) I burp out air.

***system holds pressure with cap on for hours***

If I crack the pump side banjo heading into the fluid drive section I get nice clear red fluid like cherry juice with the occasional little bubble.

If I crack the return side I get sputtering and 'milkshake' of tiny bubble foam.


I've been burping the heck out of it at the tank level.  I pump up the pressure, stop the engine and burp the air (rinse and repeat).  It always seems to have more air to burp.  The only controlling factor to the burping seems to be the amount of headspace in the tank -- less space = less burp.

I've been burping the heck out of it at the banjos.  There is always some more foam on the return side.


I've changed the height of the tank +/- 4" from my original location -- no change.

I've added a second intake hose from the bottom of the radiator to a T-barb at the pump (2x5/16) on the theory that the extra hose length is making the pump suck too hard and it's not drawing enough fluid, leading to either a starved condition or cavitation. -- still no improvement.  I'm reluctant to run it as a pressurized system simply because it's not the way it's 'supposed' to work.


I don't think it's leaking air into the system because I see no leakage and because the system holds pressure for hours after shutting the engine off.

I'm too far out of the box to see it anymore.  I need ideas or I'm going to have to return the fluid loop to oem, and I really don't want to do that.  So you hydraulic guys need to pitch me some new thoughts.

[Aaron D.]

I used to work a lot with fluid couplings-what you are experiencing I'd expect would come from low fluid level in the unit. Can you check that?

Of course the fluid couplings didn't have cooling ciruits or pumps, but besides the extra component in the converter, the principle of the thing is the same. 

[RinkRat II]

RK, possible relief valve in pump letting go too soon, sounds like pressure is the common denominator here,seems to act as internal bypassing. As you say when the pressure builds it starts to work. also what size I.D. are the hoses? :pop

    Paul :BEER:

[rodekyll]

First, thanks luap for managing to preserve recent topics.  I would have really hated to retype all that.

Hose ID are all 5/16 (8mm).  This is the natural OEM size of the banjo barbs.

Ah!  A fluid coupling guy.  Exactly the sort of expertise needed!  Thanks for checking in.  I might be coloring too far outside of the lines to be able to discuss this in purely guzzi terms.  I think the theoretical level might be more productive.

 I agree that fluid level could be a factor, although I'm not sure why.  I do not have a big reservoir or 'tank' as such.  The reserve fluid lives in the radiator.  I can see fluid circulating in the top tank of the radiator, and do not see the tops of the tubes.  Could 'low fluid level' = 'cant drop atf down the radiator tubes fast enough to feed the pump'?

I did a really mickey-mouse test last night before giving up in disgust and retreating to the comfort of wiring up my dash (comic relief).  The test was to put 500mL of fluid into a beaker and stuff all the intake and return hoses into it.  Sitting on the ground I wasn't pumping much of anything except air.  I could feel weak suction with my finger plugging the intake end.  At the return hose I had a steady stream of bubbles -- a couple dirigibles more than the system could possibly hold.  I'd occasionally get a fart of sauce at the main return, but not much flow.  The 'other' return -- the one for the shaft bearing lubrication -- was returning nothing at all.

Before we point fingers at the atf pump drive piece, let me point out that the system does show circulation when it's all connected, and it generates pressure -- up to 25# -- as long as the radiator is sealed (it might build up more, but I only proofed the radiator to 15# . . .).  If the pump shaft was stripped I'd expect a weak flutter of pressure and zero circulation.

Right now I'm thinking one or more of the following:

length of hose exceeds ability of the pump to prime
lift of hose "" ""
radiator can't drop fluid as fast as the pump wants to pick it up

Less popular things I can think of:

pump gears are worn out and can't move fluid
installed t/c mouth seal backwards

Since the system holds pressure quite well and there is no external leaking (bellhousing dry) I'm not putting much stock in a seal problem.

My next test will be to get a 1-gal gas tank with a petcock in the bottom (hard to set down!) hung from an IV stand at a level above the pump so it's all gravity primed.  I think that will tell me something, although I'm not sure what.

[RinkRat II]

 RK, after looking at a parts breakdown of the t/c drive assembly it could be leakage between the trans input shaft bushing,or the seal ring inside the flange assembly. looks like stock pressures should be 30psi so its definitely going somewhere internally.  If the converter isn't staying full that would raise the temps and aerate the fluid.  Standing By...

    Paul :BEER:

[Aaron D.]

So on thinking about this..the converter in principle doesn't need pressure to work, it would function with a fill of fluid until it overheated-the whole pump thing is primarily to get heat out of the fluid.

So it needs flow, and pressure readings are an indirect way to measure flow. And as you've shown you can have pressure without getting torque.

I think you are on the right track, there's a restriction somewhere, and the hose length is one.

Too bad you're so far away, this sounds like fun.

[dsrdave]

RK, are you using a traditional coolant type radiator for the hydraulic system ????  If so I believe that is you problem.  To cool hydraulic fluid I would use a single pass cooler such as used on an auto trans. 

[rodekyll]

Good thoughts, all!

The oem cooler is a small version of a side-tanked radiator.  If you've seen a new-fangled guzzi engine oil cooler you have a good idea of what it looks like.  The radiator I'm using is the same design except it's vertical and has a filler cap.  I had tried to find a more traditional, single-tube oil cooler but couldn't find one with any tanks -- they all simply snake a tube through fins.  Part of my design idea was combining the cooler and tank into one unit, and then hiding the unit back by the rear axle with an aux cooling fan, just in case.

Yes, the flow seems more important than pressure -- gotta keep the fluid cool.  Fotoguzzi put a gauge on his 'vert and saw much lower numbers than I do -- 2 - 12#.  I see 17 - 28#.  Both of our rigs ran well.  My other 'vert has broken its atf pump drive shaft.  It went about 2 miles before the lack of flow overheated the atf and broke it down past usefulness. 

So we have three ways to go -- literally a trivial problem:  restricted flow through the radiator, line restrictions, and a bad t/c flange seal.

I dread taking the transmission back out, and like I've said before I'm lazy, so I'm going to start by eliminating the easier stuff -- flow restrictions.  Earlier I said I was going to IV-stand a tank and gravity feed from the tank to the pump and return to whatever's handy.  That should confirm or deny the pump's ability to acquire fluid from the radiator.  I got lazier than usual yesterday and spent my trike time researching electric fluid pumps.  So I'll get with the program today before the boss fires me . . .

I've got a loosely-wrapped theory that perhaps I need to add a reservoir to the loop, more like the oem system.  I'm thinking that maybe luigi started out down the same path I've taken and found this same dragon.   

[rodekyll]

ok.

The good news is that it's not a supply side problem.
The bad news is that it's not a supply side problem

I rigged the gravity feed tank for supply and gave it about 12" drop.  I put the return hoses into big vodka bottles.  I get no difference in behavior at all.  I have gaseous pee from the t/c return and virtually nothing at the tube/bearing return.  Anyone got rum bottles?  Maybe I'll find a flavor it likes.  . .

So with the upstream end apparently working, I'm left with the pump and the fluid drive housing.  I did not mess with the timing cover/pump during this project.  I did mess with the fluid drive assembly to a fair degree.  Did I mention that the bike ran to the shop for disassembly?  I wonder what it could be . . . . .            hmmmm. .  . . . . sucking air .  . .  .    

[rodekyll]

I need all the help I can find.  Got any?   


 . . . and it's purple in my harness.  Goes to the rear view picture tube relay.  I'll see if I plumbed it into the pressure relief by accident.


Seriously though -- I've got all the components to the transmission on a bench out back.  It looks like the big seal that the mouth of the t/c spins in is the only thing that seals it.  There is an array of other seals and O-rings in the t/c output shaft tunnel, but they are all downstream of my problem and work to keep oil flowing to the bearing and shaft.  It looks conceivable that if I didn't get that t/c seal completely seated it could be blocking the bearing oil galley, which would explain the lack of return fluid via that route.  The part that I can't get my head around is how it's sucking copious air yet there is no fluid leaking -- bell housing and drain are dry.

On the other hand, I can't see how the pump could suck air either -- but one of the two is and as much as I dread checking either one out, I REALLY dread checking the wrong one first.

[fubar guzzi]

I feel your pain BRO.I am hoping youll figure it out soon because I am experiencing nearly the same symptoms.I will drink six Coors original in your honor,hope it eases the PAIN somewhat!!!

[tris]

Morning RK!

I know that I'm in danger of trying to teach my grandmother to suck eggs, but have you had a look at Gregory Benders site?

There's a big chunk on the Convert transmission (under I-Convert)

If nothing else a read might release the hidden power of your mind to identify the problem 

Tris

[Aaron D.]

Unless the converter is very weird, I suspect (still) low fluid, or possibly something wrong with the stator clutch-as in it's spinning when it shouldn't.

I think I'll peruse Mr.Bender's site.

[LaMojo]

It has to be a fluid restriction somewhere for the TC turbine to cause cavitation.

[Aaron D.]

Further to it, the fluid level is below the minimum required, or possibly the stator clutch has failed-the pressure and feed should have no bearing on torque transmission if the fluid is at the critical level.

It won't work properly if the converter is completely full, either.

[Howard R]

Without actually knowing anything about what I'm saying, it sounds to me like your "sucking air w/o fluid leaking out" situation could be due to cavitation somewhere in the system.  More likely to be a supply path blocked somewhere, because it seems that a return path blocked would lead to excessive pressure, possibly accompanied by unauthorized fluid egress.  Maybe step around to the other side of the machine and try looking at it from that direction to think of what could have been blocked during the disassembly & reincarnation process.  That's all I've got.

Good luck!

Howard

[rodekyll]

I've studied on the parts that make up the fluid drive assembly.  Assuming I got the intake and return holes in the output shaft tube properly aligned, the big seal at the mouth of the t/c looks like the only thing that could interfere with fluid flow.  Since fluid does get into the t/c and return to the tank, I think the tube is correct.  The seal could have been flopped during assembly, could be mis-seated, or could have torn.

So I'm tearing back into it today.  I've designed the trike so the bellhousing and fluid drive stuff comes out without crabbing the frame -- just the drive shaft, alternator belt, exhaust crossover and a dozen nuts and bolts need to be removed and the fluid drive section slides straight back on a rail.  I didn't want to prove it so soon.  I hope to have a stare at the seal this PM.

[twhitaker]

I've troubleshot more hydraulic and lube systems than I would care to admit. In my experience, bubbles in the system is indicative of air getting drawn in past the pump shaft seal into the suction side of the pump. I've cured a number of these by going larger on the suction line.

Your idea of eliminating the reservoir and putting it in the heat exchanger is sound.......in theory.

The flow velocity of the oil in the suction line should be below 2 feet per second. For 2 gallons per minute (SWAG)you would need to use 3/4" line.

[flangeman_70]

Hey RK,

I have seen where the internal lining of a hose is compromised/delaminates and collapses when under suction giving the effect you are seeing. If it can, air will suck past a seal but normally they just cavitatie like crazy and expel froth. I have seen this in hydraulic and fuel lines and also vacuum systems I work on in my present line of work and as we know, it is all about flow 

Food for thought

Adam

[rodekyll]

I have all new Weatherhead and Gates transmission hose installed -- 200# and 250ºf in 5/16.  It's stiff to the point of annoying and even when I try to pinch one closed with the hose clamp pliars I can't collapse it.

I'm nervous that I'm going to go through the grief of pulling the fluid drive and then discover that problem is the pump. 

[rodekyll]

I've troubleshot more hydraulic and lube systems than I would care to admit. In my experience, bubbles in the system is indicative of air getting drawn in past the pump shaft seal into the suction side of the pump. I've cured a number of these by going larger on the suction line.

Your idea of eliminating the reservoir and putting it in the heat exchanger is sound.......in theory.

The flow velocity of the oil in the suction line should be below 2 feet per second. For 2 gallons per minute (SWAG)you would need to use 3/4" line.

Any opinion on what a reasonable flow rate would be for a Convert?  I've never measured that.  I tried the other day by returning fluid to a gradient but with the bubbles and all I don't think it was a valid measurement.

Flangeman - the banjos have 8mm (5/16) barbs for all connection points.  In the past I've successfully run 1/4" return lines by boiling the ends and stretching them over the barbs.  (I only did it because ther ewas no 5/16 hose in town -- but it worked until the next time I changed hoses)

[twhitaker]

Here is the formula. You want to size your flows based on the max normal speed of the motor. I realize the ports at the pump are small but is is easier to draw through a short restriction than a long one.

http://www.hydraulicsupermarket.com/upload/db_documents_doc_16.pdf

One other thing. You say the oil cooler is mounted in back. What is its vertical relationship relative to the pump? If it is above you could consider the pump to be flooded suction. If not, then the pump has to draw the fluid. The OEM was mounted high for that reason.

Is the OEM tank vented or sealed?

[rodekyll]

The oem tank has a vent.  I have never tested to see if pressure is still possible in the oem tank.

The cooler is mounted in the back, at about the rear axle if this was still an EV.  The tank is tall.  The top nk is a couple inches above the oem tank level and the bottom is lower.  I have the pump feed at the bottom tank and the return feeds to the top tank.

With the cap off the radiator I idle at about 5# and have intermittent pressure increase as the engine revs.  I can coax it up to about 17#.  At about 12# the rear wheels begin spinning, but I can stop them easily with my hands and they don't accelerate at the same rate the engine does.  I'm 1:1 back to the differential, so I would expect some sort of proportionate increase in wheel speed with an increase in engine speed.

With the cap on the radiator the pressure comes up to about 8# at idle (initial start).  If I burp air/quickly restart I can coax the idle pressure to about 17#.  Rear wheels start spinning at about 10#.  Although there is a better correspondence of wheel speed to engine speed, it is still a 'limp' coupling -- wheels slow down and speed up at constant engine speed and I can still stop the wheels with my hands.  Returning fluid is foamy and heat builds up quickly.


cavitation.  I know cavitation happens when we spin a propeller (impeller) faster than it can move the fluid.  This cause a void on the low pressure side of the vanes and the resulting 'cavity' produces some sort of bubble.  What is the bubble made of, and can cavitation continue to generate bubbles indefinitely? 
Again -- system does not leak fluid externally.  All fittings and the bellhousing are dry.

[rodekyll]

Another troubleshooting thought --

I've been looking at 'lectric oil pumps.  It looks like I can get a continuous duty pump that operates 12vdc, handles boiling temps, lifts to 4 meters (~6.5#) and pumps a couple gallons/minute.  Does anyone see an advantage to an alternate pump source?  @~$65 it could be cheap . . .

[dsrdave]

Part of a reason for a tank in a hydraulic system is to allow the fluid to deaireate (sp) those little bubbles need to go somewhere and not back into the flow of fluid.  Just a thought.....

[rodekyll]

So you're suggesting that the tank might be for more than just storage -- it's also for settling out the bubbles.  I was hoping that the drop down the radiator tubes from top tank to bottom would act as the burping agent . . .
 . . . I can add a tank if I need to.  I mapped out some space for one when I placed stuff.  I was hoping for success with the more elegant radiator-as-tank solution.  I'm about halfway in to the disassembly for seal inspection right now.  I'll ruminate on this development as I bust knuckles. 

[dsrdave]

I'd try and rig a tank at approx. the same elevation as an OEM one and plumb it without a cooler of any kind.  See if it works.  Would be pretty easy if you have an extra OE tank laying around.

[rodekyll]

I do have an oe tank on-hand.  I'd have already done that test, but with the IV tank not helping at any level +/- (with and without the cooler in the loop) I couldn't see replumbing it back to complete oem without taking a look at that seal.  I guess part of it is that reverting to oem is a tacit admission of defeat -- that my design is bad.   I'd much rather find a brain fart in the assembly or a bad pump.  I scratched my head raw thinking this concept out and if it's possible for it to work, than yb dog ist's gonna!      problem ---> <----me

[rodekyll]

Yes, I can go back to oem if I have to.  There's nothing about the trike that prevents it except me.

I got the bellhousing off and took a peek -- everything looks pristine and normal.  Nothing visibly installed wrong.  No leaks and no signs of abnormal rubbing or stress.  When I separated the units I lost a couple ounces of fluid -- nothing to suggest the t/c is filling completely.  However fluid is dripping from the mouth of it (as it ought to with it hanging open), and it looks fresh and clean.  Where's that pounding my head on a nail smarm when I need it?

Well, unless anyone has anything for me to check while I've got it open, I think I can eliminate the fluid drive section as my problem.  I'm waiting for input before buttoning it back up.

[kirby1923]

Ok a dumb question..I don't know how the convert system is configured but I have some experience with auto type fluid converters and theory of operation from school.

So the question is this. Are you running allot more  fluid thru the system, (more than the convert specifies)  and did you bypass the cooler you installed (non oem?) and try operation?