Pressure Relief Valve question

Status
Not open for further replies.

NickZ

VIP MEMBER
Joined
Oct 31, 2018
Messages
455
Country flag
I've read several of the PRV threads but this is a little different.
While I am servicing the timing chest of my 71 Roadster, I thought I would check the PRV after I had cleaned up the body, replaced the piston & spring, and shimed it till it rattled. I thought of a simple setup that could test pressure in vs pressure out, but not sure how relevant it is because I used air pressure, not oil in the test. The setup, pictured below, is just a timing cover with the crankshaft port sealed with a cut off rubber muffler mount that I turned down to the right size so it could be circlipped in where the seal normally goes. A pressure gauge is connected to the rocker feed output and air pressure is injected from my compressor into the port where the oil pump output goes.
Pressure Relief Valve question

What I found that surprised me was that at any input pressure, there was always a loss of pressure thru the PRV, roughly 20% even before the piston moved. I double checked this by replacing the spring with a bolt so the piston can't move. With the spring, the valve would relieve more pressure as compressor pressure got above 60 psi, as I would expect a functioning PRV to do.

My question: Is this initial pressure loss, a) a function only of using air pressure, b) also present with normal operation with oil pressure, or c) because my PRV is bad?

For anyone interested, here is what I measured with the compressed air setup:
PRV testing
comp psioutput psi
2014
3023
4031
5040
6047
7054
8058
9061.5
 
Surely if the piston is ‘rattling’ ... it’s not seating ?

If it’s not seating... it’s not shutting off ?

And that’s where your initial ‘loss’ is coming from at least ?

According to Norman Whites manual the spring should be shimmed such that the piston movement is lightly restrained. In other words, shimmed such that the piston is seated.

He also says its designed to start releasing pressure above 50ft lbs. So maybe a better test is a test to ensure it’s sealing fully at anything less than that? I don’t think measuring output PSI from the valve, once it’s started lifting, is telling you much personally. The key is the pressure at the pressure side of the valve, and most key of all is that it seats when it’s supposed to and lifts when its supposed to.

Maybe you have a damaged seat or piston or similar ?

I thought dynodave did something on this? If so, I can’t find it...
 
Nick, unless I’ve misread you said you used air to test the pressure.
I would have thought that if the piston was dry you would get an air loss past it. If you soaked it in oil there may not be any loss.
Just my thoughts, I may well be wrong.
 
My question: Is this initial pressure loss, a) a function only of using air pressure, b) also present with normal operation with oil pressure, or c) because my PRV is bad?

If I remember correctly I think it has been mentioned before that testing the PRV with air is not accurate because...

What I found that surprised me was that at any input pressure, there was always a loss of pressure thru the PRV, roughly 20% even before the piston moved.

...There's no 'seating' in the valve which is why it can be set to 'rattle'. The 'restriction' is from the tight clearance between the piston and housing and the difference between hot and cold oil will alter the rate of 'leakage'.
The greater the clearance from wear, the more leakage there is likely to be even at relatively low pressures as the valve doesn't simply pop open by lifting off a seat at a certain point so some 'leakage' is to be expected.

Shimming the valve adjusts the level of pressure required to move the piston far enough to begin uncovering the oil holes in the body and this is when the valve 'opens' so the piston will begin moving before the oil reaches the relief pressure.
The question as I see it is whether there is a significant rate of leakage between the piston and valve body when the oil has reached working temperature and the pressure has dropped enough for the piston to be completely covering the oil holes.
 
I have the ss SRM valves on all my three Nortons and (touch wood!) they are behaving well in this area.
 
My question: Is this initial pressure loss, a) a function only of using air pressure, b) also present with normal operation with oil pressure, or c) because my PRV is bad?
I believe, but have not tested, that if you did your test with oil rather than air your results would be radically different. With a good PRV with two shims, I'm guessing:

1) With cold oil there would be little to no pressure loss.
2) With hot oil, a little pressure loss.

More opinions, not tested facts:

The piston has a small clearance so it can move without binding and if not constantly oiled would certainly bind. That oiling helps prevent it from leaking pressure. Notice the groove in the side of the chamber - needed to be sure the piston does not get locked by too much oil passing by. IMHO, using air invites pressure leaking and piston sticking
 
Last edited:
Notice the groove in the side of the chamber - needed to be sure the piston does not get locked by too much oil passing by.
Once the spring chamber behind the piston has filled with oil then the groove provides a vent for the trapped oil so it doesn't hydraulically lock the piston at the 'closed' position, unless that's what you meant?
 
Once the spring chamber behind the piston has filled with oil then the groove provides a vent for the trapped oil so it doesn't hydraulically lock the piston at the 'closed' position, unless that's what you meant?
Exactly!
 
Surely if the piston is ‘rattling’ ... it’s not seating ?

If it’s not seating... it’s not shutting off ?

And that’s where your initial ‘loss’ is coming from at least ?
I hear what you are saying, but I thought my test using a bolt instead of the spring would eliminate piston seating as the issue.
 
Nick, unless I’ve misread you said you used air to test the pressure.
I would have thought that if the piston was dry you would get an air loss past it. If you soaked it in oil there may not be any loss.
Just my thoughts, I may well be wrong.
I did lube the piston, but with WD40, so maybe not as good a seal as normal oil would produce.
 
The question as I see it is whether there is a significant rate of leakage between the piston and valve body when the oil has reached working temperature and the pressure has dropped enough for the piston to be completely covering the oil holes.
That is essentially my question. It would seem to me to be very inefficient if 20% of the pressure produced by the pump was being bypassed when the piston is in closed position, i.e. when pressure is below the relief pressure. So hopefully that doesn't happen with oil.
 
On all my bikes that I have oil pressure gauges on, I can see pressure as high as about 65 + PSI when cold, and that will level off to about 40- 45 PSI once warmed up at about 3500 RPMs . And it will drop lower at 1500 RPM idle when hot approx 25 PSI and less if the race bikes went out in back to back races. That to me is normal. I'm using 15-40 wt. or 20-50 Wt oils .

My Toyota Celica Supra with 5-30 Wt. when started cold will have the gauge reading quite high on the fast idle speed it runs on when cold. and yet the needle is very low when warm at idle . The factory manual states: "At normal operating temps " More than 4.3 PSI at idle and 36-71 PSI at 3,000 RPMs , So that is the way relief valves work In most cases and the Toyota engines last very well at the recommended pressures.

Warm oil flows quicker and is less thick. All I want is good flow and decent pressure. The Norton pumps do flow oil quite well as evident by the engines last a good long time.( in my case anyways)

When I worked on customers bikes , I would hook up a temporally pressure gauge to test cold and hot oil pressure, to be sure all was well and was good to go. There were defective timing cover seals being sold and heaven knows what other "mechanics" did before I worked on the bikes.
 
It would seem to me to be very inefficient if 20% of the pressure produced by the pump was being bypassed when the piston is in closed position, i.e. when pressure is below the relief pressure.
Did you seal the PRV thread as there's only a narrow band of thread (about 5 turns) between the pressure side and the vent side which oil can also leak from so no doubt air will too and perhaps more readily than oil?
 
There was a guy on this forum who said the top part of the relief valve from a pre-unit Triumph 650 fits the bottom half of the relief valve on a Commando. With pre-unit Triumphs, you should not be able to push the button in while the engine is running. I have always suspected my 850 does not have enough oil pressure. But I have been unable to find an old Triumph relief valve.
 
Did you seal the PRV thread as there's only a narrow band of thread (about 5 turns) between the pressure side and the vent side which oil can also leak from so no doubt air will too and perhaps more readily than oil?
Not for these tests, but I can try the test again with thread sealant (Loctite 565) or teflon tape.
I did test the PRV for drip thru under static pressure from the oil tank and about 11 ml of room temp 20W50 went thru the PRV in 24 hours with no thread sealant.
 
Off the topic of the testing but having to do with the PRV, the parts list catalogue shows the shim on the cap side of the spring. Should it not be on the other side of the spring inside the piston? Not meaning to interrupt the thread...
 
If they are on top, it is much easier to add or remove them while the PRV is installed on the bike. Not that you would be doing that very often, if at all.
 
Status
Not open for further replies.
Back
Top