928 Waterpump failure, warning device
10-15-2009, 02:51 PM
#1
Team Owner
Thread Starter
928 Waterpump failure, warning device
I just had a belt tension light come on in the Yellow 88 S4, so I took the belt cover off and there was a gouge on the belt, plus it was loose........
I had just retensioned the belt about 500 miles prior and everything looked great.
.I am now going to be replacing the belt and the waterpump.
I found the gouge was caused by the rollers under the crank, once the belt got loose then the rollers started hitting the belt and since they dont usually roll with the belt they were a bit tight so the gouge was caused by the roller hitting and not turning with the belt.
I removed the waterpump and found that it had been leaking a bit, but i also found that the impeller was not flush with the inside of the waterpump housing.
The top portion was about .055in. and the bottom was about .018in., a clear indication that the bearing had shifted in the housing and this is what made the belt loose.
It got me to thinking what if there was a warning device that could be incorporated into the pump to let the operator know that the bearing was letting go/ moving in the housing.
So i got some spring steel and made a couple of bends
The way this would work is the piece of metal would ride just clear of the WP pulley and if the pulley moved it would then contact the spring arm and make a squealing sound similar to the noise of a squeal on a brake disc.
If the squeal was heard then it would indicate an inspection of the waterpump and belt.
This might also be a benefit to a Porkensioner equipped car as there isnt any warning that the belt run is not functioning correctly.
IE the waterpump bearing might be moving in the housing and the auto tensioner would be taking up the slack.
I am sure that there might be another way to secure this piece of metal, I was thinking to silicone it into place, but having a bolted on part might be better.
Does anyone else have any thoughts on this idea on this????
I had just retensioned the belt about 500 miles prior and everything looked great.
.I am now going to be replacing the belt and the waterpump.
I found the gouge was caused by the rollers under the crank, once the belt got loose then the rollers started hitting the belt and since they dont usually roll with the belt they were a bit tight so the gouge was caused by the roller hitting and not turning with the belt.
I removed the waterpump and found that it had been leaking a bit, but i also found that the impeller was not flush with the inside of the waterpump housing.
The top portion was about .055in. and the bottom was about .018in., a clear indication that the bearing had shifted in the housing and this is what made the belt loose.
It got me to thinking what if there was a warning device that could be incorporated into the pump to let the operator know that the bearing was letting go/ moving in the housing.
So i got some spring steel and made a couple of bends
The way this would work is the piece of metal would ride just clear of the WP pulley and if the pulley moved it would then contact the spring arm and make a squealing sound similar to the noise of a squeal on a brake disc.
If the squeal was heard then it would indicate an inspection of the waterpump and belt.
This might also be a benefit to a Porkensioner equipped car as there isnt any warning that the belt run is not functioning correctly.
IE the waterpump bearing might be moving in the housing and the auto tensioner would be taking up the slack.
I am sure that there might be another way to secure this piece of metal, I was thinking to silicone it into place, but having a bolted on part might be better.
Does anyone else have any thoughts on this idea on this????
Last edited by Mrmerlin; 10-15-2009 at 08:47 PM.
10-15-2009, 03:09 PM
#3
Shameful Thread Killer
Rennlist Member
Rennlist Member
Secure the spring to the WP housing more betterer please. You wouldn't want the metal falling into the belt path in case it did make contact.
10-15-2009, 03:15 PM
#4
Team Owner
Thread Starter
My original thought was to silicone the spring steel onto the WP housing.
During normal operation the spring steel part wont be in contact with the pulley only if the bearing moves/fails in the housing
But thats why I posted this to get others thinking that might have better MFG ideas
Gio, if you look carefully you can see the bent spring steel part and it is shown installed at the top of the axis of the WP shaft where the pulley will be deflected if the bearing fails
During normal operation the spring steel part wont be in contact with the pulley only if the bearing moves/fails in the housing
But thats why I posted this to get others thinking that might have better MFG ideas
Gio, if you look carefully you can see the bent spring steel part and it is shown installed at the top of the axis of the WP shaft where the pulley will be deflected if the bearing fails
10-15-2009, 03:16 PM
#5
Under the Lift
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Stan:
Excellent!
You dodged a big bullet aimed right at your block and heads! So, the pulley had canted, just like with Jon's water pump, although it went a bit further in his case before it was evident to us. Good going!
I really like your warning idea. There was another thread started recently with some ideas that were more complex. Yours is simple. It would only catch the canted pulley from a bearing pack failure, but that's a serious enough issue right there and seems to be a more common failure modes recently. It would not catch migration of the impeller, but what would? It would not catch simple pulley seizure, but I suspect most seizures are not simple and probably involve what happeened to you and Jon.
Those bottom rollers need to spin freely, as you know. I may not replace the bearings on them each time, but I make sure they spin freely.
I'm not telling you anything you don't know, that this case makes the point again that any retensioning outside of the first 1500 miles (initial retension) is a red flag. And certainly, as you thought, low belt tension only 500 miles after a retension is a black flag - that belt is done and the whole system is suspect.
How many miles on the belt?
Excellent!
You dodged a big bullet aimed right at your block and heads! So, the pulley had canted, just like with Jon's water pump, although it went a bit further in his case before it was evident to us. Good going!
I really like your warning idea. There was another thread started recently with some ideas that were more complex. Yours is simple. It would only catch the canted pulley from a bearing pack failure, but that's a serious enough issue right there and seems to be a more common failure modes recently. It would not catch migration of the impeller, but what would? It would not catch simple pulley seizure, but I suspect most seizures are not simple and probably involve what happeened to you and Jon.
Those bottom rollers need to spin freely, as you know. I may not replace the bearings on them each time, but I make sure they spin freely.
I'm not telling you anything you don't know, that this case makes the point again that any retensioning outside of the first 1500 miles (initial retension) is a red flag. And certainly, as you thought, low belt tension only 500 miles after a retension is a black flag - that belt is done and the whole system is suspect.
How many miles on the belt?
10-15-2009, 03:19 PM
#6
Really neat idea.
Did you clean up those lower rollers when you did the pump last? There has been some talk about just removing them all together when doing a belt job. I leave mine in place but make sure they are smooth rolling before reassembly.
Did you clean up those lower rollers when you did the pump last? There has been some talk about just removing them all together when doing a belt job. I leave mine in place but make sure they are smooth rolling before reassembly.
Trending Topics
10-15-2009, 03:26 PM
#8
Team Owner
Thread Starter
Bill, this belt has about 2500 miles on it but i am going to replace it.
This is a rebuilt WP and it had been fine, no leaking or otherwise looseness in the shaft, and it has about 9500 miles on it.
But I backed out of the garage yesterday and the Belt light came on , I then thought.. Hummmm, I better roll it back into the garage.
So I did and this what I found.
On to the the bent metal part, I would rather make something that can be bolted on and then adjusted, in service the metal wont touch anything unless the shaft moves, and then this would be enough to warn of possible belt failure/ waterpump failure.
Also this could benefit the Porkensioner installed cars, as there isnt any warning of a belt getting loose and thus the shaft will move quite a bit without any indications of failure
Sean the rollers were new when i did the belt run back in 05 but they dont move much and with the non use I guess that they were just a bit tight, and since the belt wont usually touch them they dont usually move unless there is a belt getting loose.
Part of this idea is to keep it simple, I am sure that there are other types of metal that could also be slid onto the top of the WP that would run near the pulley
Gio that might work but i was thinking to keep the warning part away from the belt and to place it where the pulley would be going if it was failing ,.
Once the bearing moves the shaft tilts up and with this piece of metal slightly rubbing on the shaft it should make a squealing noise similar to a brake pad wear sensor
This is a rebuilt WP and it had been fine, no leaking or otherwise looseness in the shaft, and it has about 9500 miles on it.
But I backed out of the garage yesterday and the Belt light came on , I then thought.. Hummmm, I better roll it back into the garage.
So I did and this what I found.
On to the the bent metal part, I would rather make something that can be bolted on and then adjusted, in service the metal wont touch anything unless the shaft moves, and then this would be enough to warn of possible belt failure/ waterpump failure.
Also this could benefit the Porkensioner installed cars, as there isnt any warning of a belt getting loose and thus the shaft will move quite a bit without any indications of failure
Sean the rollers were new when i did the belt run back in 05 but they dont move much and with the non use I guess that they were just a bit tight, and since the belt wont usually touch them they dont usually move unless there is a belt getting loose.
Part of this idea is to keep it simple, I am sure that there are other types of metal that could also be slid onto the top of the WP that would run near the pulley
Gio that might work but i was thinking to keep the warning part away from the belt and to place it where the pulley would be going if it was failing ,.
Once the bearing moves the shaft tilts up and with this piece of metal slightly rubbing on the shaft it should make a squealing noise similar to a brake pad wear sensor
Last edited by Mrmerlin; 10-15-2009 at 03:41 PM.
10-15-2009, 03:27 PM
#9
Under the Lift
Lifetime Rennlist
Member
Lifetime Rennlist
Member
This incident seems like it makes another point....
Had you installed a Porkensioner, you would not have known about the low tension caused by the pump pulley canting or other cause until it was possibly too late.
Oh, Stan, I now see you just mentioned this. Yes, your warning tang would help, but a better solution would be something like this combined with low tension warning that would pick up other causes. This should go to another thread, but I'd like to see the Porkensioner have a warning when the Audi tensioner shaft over-extends.
Had you installed a Porkensioner, you would not have known about the low tension caused by the pump pulley canting or other cause until it was possibly too late.
Oh, Stan, I now see you just mentioned this. Yes, your warning tang would help, but a better solution would be something like this combined with low tension warning that would pick up other causes. This should go to another thread, but I'd like to see the Porkensioner have a warning when the Audi tensioner shaft over-extends.
10-15-2009, 03:34 PM
#10
Rennlist Member
Join Date: Jan 2006
Location: central cal
Posts: 975
Likes: 0
Received 0 Likes
on
0 Posts
Stan good catch and an interesting idea. You might be able to bolt something on to another spot and have it bent in a way that it contacts like yours, but can't come off...
I'm not in favor of it though, because that will prevent Greg me and from getting rich off the block damage repair system....
Steve
I'm not in favor of it though, because that will prevent Greg me and from getting rich off the block damage repair system....
Steve
10-15-2009, 03:56 PM
#12
Team Owner
Thread Starter
Steve i cant help but think if Greg might have had one of these little tell tales on the failed pump on his Big engine, it might have prevented a lot of damage, and since a WP pulley fails upwards it would usually give a warning
10-15-2009, 03:58 PM
#13
Rennlist Member
I like the idea of having a warning. Usually the brake "squeal" indicators have the metal strip perpendicular to the rotor surface to insure that there is vibration/squeal when contact is made. Without testing I would hesitate to recommend this as a reliable warning -- no guarantee the driver would hear it.
Why not take a brake pad sensor, sand it down until the wire is just visible(so you know where it is), then coat it with some clear paint(clear so you can see the loop and line it up for installation, paint to prevent the wire from corroding). Mount this with minimal clearance to the pulley. You would have to include a fast-response latching relay or electronic circuit to ensure that momentary contact would give you a steady warning, especially if you want to use the original warning circuit that includes a delay. BTW Gio, any "alert on contact" system has to be in the top quadrant of the pulley. The position you show would have to be in constant contact.
Maybe something similar could be done for the pulley on the Porkensioner, or a sealed microswitch could detect movement since in that case more movement is acceptable due to the nature of the setup, i.e. the belt can stretch quite a bit before tension is compromised compared to the OE tensioners. This could also be wired to the original circuit, both warning switches wired in parallel.
It might make sense to either modify the original central warning circuit on the CW PCB to eliminate the delay or to wire a separate circuit. IIRC the original warning system waits 3 minutes after a fault is detected before it illuminates, and 3 minutes can be an eternity. If this is implemented in an IC, it probably can't be changed. If it's a simple RC timer on the PCB it could be eliminated easily.
Why not take a brake pad sensor, sand it down until the wire is just visible(so you know where it is), then coat it with some clear paint(clear so you can see the loop and line it up for installation, paint to prevent the wire from corroding). Mount this with minimal clearance to the pulley. You would have to include a fast-response latching relay or electronic circuit to ensure that momentary contact would give you a steady warning, especially if you want to use the original warning circuit that includes a delay. BTW Gio, any "alert on contact" system has to be in the top quadrant of the pulley. The position you show would have to be in constant contact.
Maybe something similar could be done for the pulley on the Porkensioner, or a sealed microswitch could detect movement since in that case more movement is acceptable due to the nature of the setup, i.e. the belt can stretch quite a bit before tension is compromised compared to the OE tensioners. This could also be wired to the original circuit, both warning switches wired in parallel.
It might make sense to either modify the original central warning circuit on the CW PCB to eliminate the delay or to wire a separate circuit. IIRC the original warning system waits 3 minutes after a fault is detected before it illuminates, and 3 minutes can be an eternity. If this is implemented in an IC, it probably can't be changed. If it's a simple RC timer on the PCB it could be eliminated easily.
10-15-2009, 04:08 PM
#15
Nordschleife Master
Stan, I like the idea.
How about making a bar out of a non conducting material such as a nylon or delryn etc. The bolting it to the two M6 holes at the 4 and 10 ok positions. Then riviting a bar of steel down just not touching the pulley. Then you can attach a wire to that bar and wire it in to the factory warning light.
However the factory warning light only turns on when the ground disconnects. So one would need to open and modify the central warning box under the dead pedal to give the right reading. Or use a 5 pin relay to switch the signal from a missing ground to a broken ground.
How about making a bar out of a non conducting material such as a nylon or delryn etc. The bolting it to the two M6 holes at the 4 and 10 ok positions. Then riviting a bar of steel down just not touching the pulley. Then you can attach a wire to that bar and wire it in to the factory warning light.
However the factory warning light only turns on when the ground disconnects. So one would need to open and modify the central warning box under the dead pedal to give the right reading. Or use a 5 pin relay to switch the signal from a missing ground to a broken ground.