hufflepuff

2005 6MT being used for HPDE, TT, and autocross. I think I'm having issues with my aftermarket pulley kit, which replaces the alternator and water pump pulleys but retains the OEM crank pulley.

My previous engine failed due to a damaged front bearing that the rebuilder claimed was partially due to over-tensioned accessory belts. Okay, now I'm afraid to repeat that scenario and I've been trying to run them as loose as possible.

However, I think my water pump belt is TOO loose, which is causing the pulley to slip and the car will begin to overheat. I can reach in and turn the water pump pulley with my hand. So I tighten it up, and the issue is reduced. BUT, the next day with the engine completely cold, the belt is loose again. I'm not sure if the belts are stretching, or I'm not torqueing down on the lock bolts hard enough.


One issue with the OEM setup geometry is that the belt acts over a narrow angle on the H2O pulley, reducing surface area and friction. The alternator has a ton of angle and friction...


So, some questions:

1. Has anyone heard of causing a front bearing failure due to over-tightened accessory belts? Should I be trying to keep them loose to "prevent" this type of failure?

2. Should I be able to spin the water pump pulley with my hand?

3. Has anyone else noticed that the belt tension goes from low when cold to much higher/tighter when the engine is at operating temperature? Should I be setting the cold or hot tension when following mazda FSM recommendations? If I set the tension when cold, it becomes quite tight (but maybe it's supposed to be?)

hufflepuff

just as an update to my own thread after many dozens of track days:

1. No failure yet although i still keep tension as low as possible without slipping
2. No, you really shouldn't. maybe just barely able to spin by hand is acceptable tension on the low end.
3. I just accept this as normal now.

Loki

iTrader: (1)

The belts can be looser when cold, sure, but not hugely so. ItsI not so much the tension as cold and hot friction coefficients. If the aftermarket pulleys are anodized and nicely polished, that might be reducing friction below the OEM design.


Adjust the belts cold. If issue persists, swap back to OEM water pump pulley. See if that does it.

I'm sure that if you ratchet the belts up to 50Nm, that will affect the bearing, but as long as you're close to spec tension, it's fine. They don't need to be extra loose at all.

NotAPreppie

I would expect the belts to be tighter when cold since rubber (as a general category) have about 7x the thermal expansion coefficient of steel.
https://www.engineeringtoolbox.com/l...ents-d_95.html

Maybe the rubber in the belt is just hardened when cold and the coefficient of friction is low? No idea how to explain what you're seeing.

Get yourself a basic force gauge that can go up to at least 22lb of force. The type they use to measure the force require to open ADA-compliant doors might work. A hanging weight scale should also work.
If you get it adjusted correctly and still have an issue then we know it's something other than tension.

Loki

iTrader: (1)

^^ this IMO. It makes sense, you know how when sometimes you hear belt squeal in the winter when someone starts their winter beater in the morning? Pretty sure it's the same effect. Plus aftermarket pulleys that may or may not hug the belt ribs correctly and the oblique belt angle at the water pump pulley. The circumstances are all there.

Plus the pulleys are aluminium, so only about 3 times less thermal expansion than pure rubber... and the belt is reinforced anyway, so it's not just rubber.

If the tension is set correctly, I'd just replace the pulley back to OEM. It's not doing a heck of a lot anyway.

NotAPreppie

That or do something to improve the coefficient of friction of the pulley itself. Not sure what that would be or how it would impact the life of the belt but it's an avenue of exploration if you've got the belt tensioned correctly.

bd32322

Sorry for reviving an old thread - but how do you check whether the deflection on the belts is within spec. How much force should be applied to the points shown in the diagram in the first post?

Thanks