Lost a water pump belt - some data & lessons learned

[FABombjoy]

Had an exciting moment on the highway this weekend when the water pump / alternator belt snapped. Everything turned out OK, no engine damage, only spoiled plans. Belt was a Dayco top-cog, no more than 5 years old. I did recover the belt which was flung up into the engine compartment and it had clearly just plain wore out, rather surprising given the age and limited mileage. Has numerous cracks in the rubber along the top side.

Megasquirt 3 is always recording to SD card which gives a glimpse of how quickly things go down (or up, rather). Figured a few people might find this interesting:

Ambient temp: About 85F, sunny, humid, AC on full.

From the moment of belt loss (battery suddenly drops 2v, the exact point is pretty obvious) there was no measured temp increase for 30 seconds at the Y-pipe.
After 30s, temperature increase 1.5 degrees F per second. I'm guessing that it took a little time for coolant to circulate by convection which is why there was such a long delay.
Dash gauge read 260 when I was finally safe to engine-off.
I boiled away about 1.5gal of coolant/water in 80 seconds.
I'm sure the charge light came on but I didn't notice it, being red and right about behind where my hand is on the wheel. I did notice the abnormal gauges (high temp, low volts) but the 2nd half of the game had already begun.

I carry a tool kit & spare belt, but had never actually used those tools to change a belt. A magic "all in one" type wrench that seemed useful on paper could not fit between the engine bay wall and the bolts locking the AC belt tensioner. I cut the AC belt off so the water pump / alternator belt could be installed which was not ideal. A stock-style alternator tensioner is not an easy affair to work when you're on the shoulder, and the shoulder quickly transitions to a drainage ditch. Once cooler, was able to drive the car in surges to a weigh station 1 mile up the road and at least get off of the shoudler. Given more time I may have been able to work out a solution to release the AC belt but traffic passing at 80MPH only a few feet away, plus I had one of my kids with me, and the problem called for a more rapid solution.

Lessons learned:
-If you lose this belt you have at minimum 30 seconds to "find a landing spot"
-250F+ is a scary number to see but absolutely no damage resulted
-Water boiling stirs up crud, consider a full-flush if this happens (currently in progress)
-If you carry spare parts, check that your tools are actually useful to install them :D
-Consider a turnbuckle type alternator tensioner instead of stock

A stranger stopped to help, assuming it was a fire given the cloud pouring from the engine bay and came running with a fire extinguisher. He also drove up to the nearest exit to bring back water. Good people are out there!

[David T]

One of the top reasons a belt will fail is if the pulleys are not lined up, ie, in the same plane. Use a string or straightedge and make sure the alternator is lined up. Other reasons for a belt to fail include;
A bent pulley
Stretching a belt to put it on (causes internal damage to the cords)
Overtensioning or undertensioning the belt. There is not so much load on the belt that is should jump off unless the water pump bearings seize up or the fan on the alternator chews up the side of the belt.

[Bitsyncmaster]

So what voltage did it instantly fall to. I set my electronic otterstat to beep below 12.5 volts to warn of a belt break.

[FABombjoy]

First 0.7s: 13.5 to 12.4
Next 8s: drops to 11.92
--From this point it slowly tapers down as battery slowly depletes
At 80s: 11.68

I have a CAN bus gauge with programmable warnings. I'll be setting it up to trigger a "Thrown Belt" alarm, something like RPM > 700 && BattV < 12.5

[FABombjoy]

https://i.imgur.com/MGHhYQl.png

[Bitsyncmaster]

First 0.7s: 13.5 to 12.4
Next 8s: drops to 11.92
--From this point it slowly tapers down as battery slowly depletes
At 80s: 11.68

I have a CAN bus gauge with programmable warnings. I'll be setting it up to trigger a "Thrown Belt" alarm, something like RPM > 700 && BattV < 12.5

Sounds good. My otterstat is powered by the RPM relay so it only warns when the engine is running.

[Helirich]

So did you have an extra AC belt too?

I guess you know you can put a belt on with a screw driver. Not perticularly fun or safe. But I seen guys do it many times. I've done it on a VW, but it's much easier/safer on them because of the access.

[FABombjoy]

So did you have an extra AC belt too?
I do now, as of 1PM LOL

Finding quality belts in-stock for a reasonable price was a challenge. O'Reillys wants something like $30 for each belt. The local NAPA had them and they're made by Gates, XL style. The cheap one is a non-XL and was only maybe $3 or 4? Cheap enough to keep around for an extra.

I have some extra plumbing above the muffler for the turbo, as well as a non-stock config on the LH water pipe side which does make things a bit more challenging. The AC idler pulley tensioner bolt has been hacksawed off to make room for the turbo oil/water lines, but a tire wrench is enough to leverage that belt tension well enough. This is actually how all of the Island TT kits are set up.

I've seen the screwdriver trick but have the parts to make a double-heim alternator tensioner on the way. Managed to pick up a surplus universal part for about $15 that should work. I probably could have levered the AC belt off using the screwdriver but would prefer to hone that technique under better circumstances.

[FABombjoy]

Plan "C" is ready for the next belt failure.

https://i.imgur.com/pZ4YK2b.jpg

[Helirich]

That's a pretty cool rack. Does it go up with the louvers or do you have to remove the four knobs? If it stays will the struts hold it up? (Nothing on it) Do the pieces that the knobs screw into remove easy or do they stay when not useing it? Does the rack fit into the trunk?

Nice bike also.

[FABombjoy]

It's the OEM optional rack - I would rate it "ok". Has to come off to open the louvers. Fits in the trunk but a little awkwardly, the OEM kit comes with a hold-down clamp to keep it from rolling around (much). The hardware that mounts to the quarter panel has crappy paint and is prone to rust, and you have to cut the "support buffers" creatively to make it all come together. But what it lacks in finish it makes up for in pragmatism.

[FABombjoy]

The great 2021 overheat event just keeps on giving!

Less than a week before the Michigan fall drive I started to see coolant running down the front of the motor. USB borescope camera showed that the weep hole was dry but found at least one hose clamp dripping. Excavated down in to the valley and found 6 loosened water pump & heater hose clamps and a flooded valley. These hoses & clamps were installed in 2018 and have had numerous leak-free trips the following years. I'm thinking the very high temp cycle cause them to sink in to the hoses and the clamps lost tension.

Tightened a few clamps, doubled-up a few, and swapped some thin band Norma clamps for thicker/wider Ideal Smartseal. Successfully held pressure for many hours and was all set to drive a few days before departure. Warmed up the car in the garage, went for a little drive, all is good.

Friday morning, 20 minutes out of town, coolant temp is holding rock solid at 145F. A tiny voice in my head reminds me of the "fail safe" thermostat also installed in 2018. FWIW the "overheat lock-open" feature does work, quite well.

https://i.imgur.com/MhFmAhT.png

Because more data:
-A new functioning thermostat gained 0.21 deg/s F, compared to a stuck open which gained 0.155 deg/s.

[Helirich]

Because more data:
-A new functioning thermostat gained 0.21 deg/s F, compared to a stuck open which gained 0.155 deg/s.

I'm a little lost here. When you say "gained" Gained from what? The 190 where it opens? Or are the decimal points in the wrong place? (21 degrees instead of .21) Earlier, you said the stuck open gave 145 degrees. (Not 155)

Ive never heard of the Fail safe thermostat. I assume from your description, it latches in an overheat situation. I guess that means you must unlatch it manually? (Take it out and unlatch) Or is it permanent open after that? What is the logic behind that?

[FABombjoy]

I'm a little lost here. When you say "gained" Gained from what?
From cold, for every second of operation, a functioning tstat increased engine temps 0.21 degF compared to the 25% slower stuck-open rate of 0.155 degF

From the company:
https://motorad.com/tech-tips/fail-safe-thermostats/

I'm not sure if it's worth the effort. But it was one of the few options at the time with an integrated jiggle valve.

[dmcman73]

The great 2021 overheat event just keeps on giving!

Less than a week before the Michigan fall drive I started to see coolant running down the front of the motor. USB borescope camera showed that the weep hole was dry but found at least one hose clamp dripping. Excavated down in to the valley and found 6 loosened water pump & heater hose clamps and a flooded valley. These hoses & clamps were installed in 2018 and have had numerous leak-free trips the following years. I'm thinking the very high temp cycle cause them to sink in to the hoses and the clamps lost tension.

Tightened a few clamps, doubled-up a few, and swapped some thin band Norma clamps for thicker/wider Ideal Smartseal. Successfully held pressure for many hours and was all set to drive a few days before departure. Warmed up the car in the garage, went for a little drive, all is good.

Friday morning, 20 minutes out of town, coolant temp is holding rock solid at 145F. A tiny voice in my head reminds me of the "fail safe" thermostat also installed in 2018. FWIW the "overheat lock-open" feature does work, quite well.

https://i.imgur.com/MhFmAhT.png

Because more data:
-A new functioning thermostat gained 0.21 deg/s F, compared to a stuck open which gained 0.155 deg/s.

Were the hoses silicone? Clamps on Silicone hoses need to be tightened every now and then. For the hoses on the back of the water pump, it's usually recommended to use just standard rubber hoses to avoid that issue.

[FABombjoy]

Y-pipe hoses are silicone but the rest were not. Will switch back to quality rubber hose at the next water pump service though as I don't see the benefit and they can be more difficult to seal. The 45-degree hoses at the heads were rubber, one of which had developed a regular drip (Norma clamp). You really shouldn't have to tighten clamps repeatedly if the correct types are used, rather follow the initial torque setting, followed by a waiting period, then a re-torque. If you have to keep torquing then the hose is yielding or the clamp is the wrong size or deforming.

Overall I'm not sold on Norma clamps in hard to reach areas or for for silicone use. They're nice as they don't expose the hose to serrations and seem built solidly, but they are not particularly wide. I have higher hopes for the SmartSeal which are almost twice as wide w/ a protective band to prevent serrations from digging in. The system held 18psi cold once everything was reclamped and such and survived last weekends 500+ mile journey leak free. Fortunately I didn't need the heater but did temporarily reprogram a few of the warm-up temperature points to get some closed loop benefits.

[Helirich]

From cold, for every second of operation, a functioning tstat increased engine temps 0.21 degF compared to the 25% slower stuck-open rate of 0.155 degF

From the company:
https://motorad.com/tech-tips/fail-safe-thermostats/

I'm not sure if it's worth the effort. But it was one of the few options at the time with an integrated jiggle valve.

That web site says that standard thermostats lock closed when your car over heats. Why would that be? First I've heard of that. I've overheated lots of times and never changed the thermostat on my truck. It darn sure did not lock closed. I'm not sure I want it to lock open either. Why wouldn't you just want it to work as normal?

[Ron]

That web site says that standard thermostats lock closed when your car over heats. Why would that be? First I've heard of that. I've overheated lots of times and never changed the thermostat on my truck. It darn sure did not lock closed. I'm not sure I want it to lock open either. Why wouldn't you just want it to work as normal?It's an advertizment...! Word games.
A standard thermostat doesn't "lock" shut -- They can get "stuck" shut and cause an overheat (due to rust, debris, wax degrading, etc). (They can stick in any position.) Normally they do as you suggest, just work, no locking to it.
Check out the pic HERE (https://www.autozone.com/cooling-heating-and-climate-control/thermostat/p/failsafe-thermostat-7200-180/139156_0_0). When the "fail-safe" thermostats reach an overheat temp, the tabs catch, preventing it from closing (unless you manually release them). (They don't actually revert to a safe condition, because the engine will run too cold. Nor do they prevent overheating in the first place. So they are not really "fail safe".)

[dmcman73]

Y-pipe hoses are silicone but the rest were not. Will switch back to quality rubber hose at the next water pump service though as I don't see the benefit and they can be more difficult to seal. The 45-degree hoses at the heads were rubber, one of which had developed a regular drip (Norma clamp). You really shouldn't have to tighten clamps repeatedly if the correct types are used, rather follow the initial torque setting, followed by a waiting period, then a re-torque. If you have to keep torquing then the hose is yielding or the clamp is the wrong size or deforming.

Overall I'm not sold on Norma clamps in hard to reach areas or for for silicone use. They're nice as they don't expose the hose to serrations and seem built solidly, but they are not particularly wide. I have higher hopes for the SmartSeal which are almost twice as wide w/ a protective band to prevent serrations from digging in. The system held 18psi cold once everything was reclamped and such and survived last weekends 500+ mile journey leak free. Fortunately I didn't need the heater but did temporarily reprogram a few of the warm-up temperature points to get some closed loop benefits.

I haven't tried them (yet) and maybe someone else on this board has, but what about spring clamps? They are supposed to provide constant pressure on the hose no matter what. Granted they are a PIA to remove and install, but may be worth it for the hoses in the VOD? I was thinking of switching to these on my next coolant flush/change.

[FABombjoy]

I haven't tried them (yet) and maybe someone else on this board has, but what about spring clamps?
There are a few options in constant torque / spring loaded / T-bolt clamps. I was a little time limited otherwise I might have explored them. Those can be pretty big and I don't think they would fit all of the water pump fittings with the larger adjusters. Napa did have CT type clamps in stock for $10 each which was a bit more than I wanted to spend. A pair of the Smartseal clamps was around $6? And I knew they'd fit all of the hoses involved.

Basic T-bolt clamps might work too but they have to be closed to the correct sizing as their adjustment range is limited. The market is flooded with cheap T-bolt clamps which are probably OK for air use, but for fluids would only select name brand parts.

For my next T-stat I'll probably avoid the fail safe type. For no other reason than I have all sorts of warnings programmed in now related to coolant temp that will be difficult to ignore. Maybe there is a failure mode that damages the mechanism or wax pellet somehow, or if it pushes in to the t-stat housing and bends. If so I can appreciate a design that says "the valve has opened much further than the normal position, so assume something has failed and at least prevent the t-stat from exacerbating the condition".