I have the failed pulley, and it is not failed around the hub, looks fine actually, if you mean failure as in plastic deformation or crack initiation. Lets say the hub deforms and it looses the press it, why wouldn't a weld on the end hold it still hold it since the axial force should be relatively small?

Can you try 2x380lb=760lb of force just for ***** and giggles?

 

https://www.performancetrucks.net/fo...9/#post4742050

I would think that if you get it to stay on, you still wouldn't reduce stresses around the hub and it would end up broken like in the post above. I think it needs a combination of holding it on the shaft and strengthening the pulley to reduce the stresses around the hub to an acceptable level. 50% of yield is about the endurance limit for steel so bringing that down to about 30% would avoid failure from fatigue and then you just have to worry about holding it onto the shaft.

The weld will likely cause embrittlement in the heat affected zone in the shaft. 309 may help, but it would be worth seeing if the weld is cracking or if the shaft is cracking close to the weld. If the shaft cracks close to the weld, I wouldn't the filler won't make much of a difference. Just some food for thought.

I need to get back to work, but this is so much more interesting at the moment.

edit-photobucket isn't cooperating I attached screenshots. One is the stress contour map, then displacement contour map, and an exaggerated deformation one.

 

Just saw the last sentence in your previous post. Let me try to get that done really quickly.

I should also note this is just to see the deformation in the bigger diameter part of the pulley and the stresses induced around the hub. The way the constraints are setup you can't really see the deformation that would occur in the hub.

EDIt- found an error in my calculation when I was converting to pressure. I was applying 75% of the actual load. I will redo it with the real load and the 760lbs

 

Ok, I got the analysis done with both the full 480 lbs and the 760 lbs too.

With 480 lbs, the stress around the hub is about 115 MPa and max deformation is 0.15mm at the belt landing. I changed my constraints so I could see some of the deflection in the hub and it is deforming, I'll see about finding the magnitude at the hub if you're interested.

With 760 lbs, the stress jumps to 180 MPa around the hub and max deformation to about 0.244 mm. So, even more now I think that if you can get the pulley to not fall off it will break. So, I don't think the pulley is made from the weaker hot rolled mild steel (1010). It's probably something like normal A36 with a yield strength of about 250 MPa, I would be surprised if it was anything stronger.

There's room for improvement on the pulley I think. It would be interesting to do a quick bending stress calculation on the shaft to see what kind of stresses it would see in these extreme situations and stock. This all makes me want to take apart the old water pump I took off my 5.3 to see what can be done to swap the shaft and pulley.

 

There are 2 different styles of the waterpump, most obviously because of the different style thermostat (1 piece 99-03 vs 2 piece 04+). There is another difference on the back side of the pump; the early style is bolted in with 5 or so bolts around the perimeter and the newer style is pressed in. I broke one of the old ones so I might take it apart tonight and see what it looks like on the inside. I have a hard time thinking the shaft is bending under the load.

What was the amount of wrap you used in the analysis? It looks like 160 degrees or so where I think the real wrap amount is more like 240 degrees.

I thought about putting a nut on the snout and welding it to the pulley on the land area to strength that area, but the thread size is very odd. I measured it at M29x1.0 which doesn't really exist anywhere, even M30x1.0 is basically impossible to find. So anyone have an old fan clutch they want to donate? Plan is to cut off the nut and make that work....

Other option would be a custom pulley, but I have a feeling that would be almost as much as the fancy PRW pump.

 

I've been following along, and I was hoping you would find a solution. Your last comment about the cap welded on is an interesting one that may work to stiffen up the area and prevent the deformation around the shaft. I know someone used to make beauty caps for the pumps, but I can't find anything when I search about them.

 

Yea I spent this morning trying to find them but cant. I almost bought one at one point too. If I knew for certain what size the threads are I would just order a big nut for it. M29x1 or M30x1 just doesn't seem right. I may take the pump into Fastenal or something and try stuff until it works.

 

The wrap angle is 180 degrees, if I go any more than that the components of the belt pressure facing each other cancel out and we only get the components pulling the pulley in the resultant direction the belt does. So I just went 180, if anything this is being slightly conservative which I tend to do.

No thread is too odd with single point threading. I was also considering a piece that threads on and can be welded at the front of the hub and on the pulley wherever it contacts. The beauty caps I remember seeing were aluminum. Maybe the fan clutch description has the thread listed? Yeah, once we talk custom pulleys the PRW pump may be cheaper/easier. What's the fun in that?

 

Atomic, do you need the mechanical fan that screws on to the water pump? If so I have a couple laying around from e-fan swaps that I have no use for.

 

Wow, some real engineering going on in here.

The nuts were made by T-Mac sold thru LC32. According to Tom, the thread was M30x1.5.
Gaging interest-Decorative water pump pulley nut. - Chevy Trailblazer SS Forum