Mercruiser 470 Water Pump Seal Leak

[Argh!]

Last winter I rebuilt my Mercruiser 470, sent the cam to a cam shop where they welded up the groves where the water pump seal rides and ground it back down. Not knowing if they got the hardness right I put a speedi sleeve on it just to be safe. After one or two tanks of fuel I was getting coolant out the weep hole. I followed the procedure documented at Mercruiser 470 Coolant Leak using the in-boat method. Since the cam / speedi sleeve looked perfect I just replaced the seal. After using about a tank of fuel the leak returned.

So I read up on it some more, made sure I was not putting the seal in backwards, not mis-aligned with the speedi-sleeve, and generally tried to be more careful (but not finding anything I could point to as the root cause). I thought maybe the third time was the charm because after a full tank of fuel it was not leaking. But half way through the second tank the leak was back, added about 6 oz. of coolant, and by the end of the second tank my coolant was dangerously low.

I have it apart again and can't figure out what is going on. The seal looks fine as well as the speedi sleeve but the only way to get coolant out the weep hole is through the seal. Does anyone have any advice?

 

[stonyloam]

Re: Mercruiser 470 Water Pump Seal Leak

Did you apply sealant to the outside of the seals when you installed them? The manual recommends Loctite 8831. Could be leaking around the outside of the seal. Just a double check on the seal direction- Garter spring side away from the block, (both) toward the pressure (coolant) side.

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

Thank you for the reply. Sorry I was slow getting back to you, I thought I would get an email if someone replied to my post, I will look for an account setting for that.

I double and tripple checked that I was putting the seal in correctly, with the sping side exposed to the pressure (coolant). I wish that could be the problem, it would be an easy fix. Also, I did use locktite all three times I put this thing together. It is a tight fit and the hole in the cover for the seal is in good condition, I really don't believe that it can leak at close to a gallon an hour around the outside of the seal three times in a row given the condition of the parts and when using the correct procedure to assemble it. Also, if it were leaking around the outside of the seal it would be strange that it would start out not leaking and then start leaking at an increasing rate since there is no movement between the seal and the hole into which it is pressed.

I am really stumped by this one.

 

[stonyloam]

Re: Mercruiser 470 Water Pump Seal Leak

Me too! Did you use OEM seals? See what condition the lip is in when you pull it out, it might be a defective seal.

 

[Eric 3.7LX Alpha 1]

Re: Mercruiser 470 Water Pump Seal Leak

I wonder if it's possible the front cover itself has a pinhole allowing the coolant to drip, and get around the seals?

 

[stonyloam]

Re: Mercruiser 470 Water Pump Seal Leak

That is a good thought Eric, but if the coolant is coming out of the weep hole there is a very limited area where that hole (or crack) could be. The weep hole leads directly to the small chamber between the seals, so there would have to be a crack in the casting that holds the seals that leads directly to that chamber, or directly to the weep hole which is certainly a possibility. You know, that has me thinking that the casting itself may be warped so the seals are not seating properly, the covers are fairly fragile (don't ask me how I know LOL). There is a brand new front cover for $50 on e-bay, might be worth the investment. Good luck.

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

I am thinking the same that the cover could be warped. I will try to grab that cover off of ebay and try it. I inspected the cover for cracks in direct sunlight and I don't see anything. Plus all three time it did not leak until the second trip out. If it was cracked it should leak right away. The cover looks straight but that is not very precise information.

Here are a few pics of the parts. I will try to get some measurements with the cover on and seals out to see if I tell if it is warped.

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

I just purchased the new cover from ebay.

 

[EddiePetty]

Re: Mercruiser 470 Water Pump Seal Leak

....two thoughts.

Are you lowering the oil pan to allow alignment of the pump cover over the dowels or are you forcing the cover at an angle?

I recently did the sleeve job on a 470 and I swear the forward sleeve (pictures 3-5) was closer to the aft sleeve.....and BTW, what are the two wear rings BETWEEN the sleeve in the pictures?

 

[stonyloam]

Re: Mercruiser 470 Water Pump Seal Leak

A lot of the guys use two spedi sleeves to make sure the seals are riding on the spedi sleeve surface. Could it be that your forward seal is too close to the front edge of the sleeve and getting torn up?

 

[RenkenII]

Re: Mercruiser 470 Water Pump Seal Leak

It looks like the seal farthest from the block is wearing fast or getting hot? I think it may be too tight and heating up, causing the seal to fail? I just did this repair too and "knock on wood" I'm in good shape so far.

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

Eddie, I was not lowering the oil pan because I would likely have to pull the engine to get to all the bolts. I did my best to keep the cover as straight as I could as it went on and the RTV I used worked as a lubricant as I forced the oil pan gasket across the oil pan. I agree its not ideal but I think it went on at a reasonable angle.

The first time I put it together I tried the three seal method and that caused the extra set of wear rings. Of course the weep hole should not be blocked off so I drilled a hole in the outer race of the seal to provide a path for leaked coolant. Since the spring side of the seal faces the coolant I don't think the added seal really is doing anything except for adding extra protection before the coolant makes it to the oil. So after the first time I just used two seals.

Terry, the forward seal is easy to see before installing the water pump impeller and about 0.030" of the sleeve was visible. The failed seal did not feel torn when I ran my finger around it.

RenkenII, I wondered the same thing but it should be totally surrounded by coolant. With that said the coolant got really low on my last trip to shore so maybe it did get hot but probably after I lost significant coolant. I might be replacing a head gasket soon too, old boats are so much entertainment!

I am using the sleeves because I don't know if the cam shop got the hardness of the repair to OEM spec (where they welded up the grooves). I am on the fence as far as pulling the sleeves off is concerned. It is hard to weigh if or how much they might be contributing to the current problem and if or how much they might me delaying a future problem (eventually more grooves). I can pull the sleeves but if they are not the root cause it will probably leak even sooner and faster. I really don't want to do this job yet another time.

The seal does not seem to be damaged. If the cover is not warped then maybe the cam shop dropped the cam or something. Maybe I should get a dial indicator and check the cam for run-out.

 

[EddiePetty]

Re: Mercruiser 470 Water Pump Seal Leak

...a couple of continuing thoughts.

The oil pan need not be completely dropped. Simply loosen the bolts sufficiently to allow the forward end to drop about a quarter inch...the aft-most bolts need not loosened. Without lowering the pan you run the risk of damaging the seal lip as it is forced at an angle over the can nose.

Provided the machine shop returned the cam nose to original specifications (diameter AND true), I see no need to be concerned over hardness as this is a seal journal, not a bearing journal. Adding the sleeves is thus overkill and maybe leading to your problems

It is interesting that you report no problems with the oil seal. Perhaps it's relation to oil is saving it. Soooooo, perhaps the water seal is lacking lubrication....are you running antifreeze or tap water?


I still believe the water seals sleeve is too far forward....examine Renkenll's pictures above and compare to your installation.

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

I will look into dropping the pan. If I remove the heat exchanger I might be able to reach enough of the bolts. I really don't think I damaged the seal during assembly but I am not taking any chances at this point.

The typical failure mode of the cam nose is that grooves are worn by the seal leading to leaking. A harder material should wear more slowly but in this case the wear process involves corrosion so perhaps you are right that the hardness is not so important. So yes, I have been worried all along that the sleeves are overkill and could be leading to the problem. On the other hand many others have run the sleeves successfully. Still, I think this time around I am going to try taking them off.

Why would you suspect oil seal trouble? I am running antifreeze at a 50/50 mix.

My seal is centered perfectly on the sleeve. I think the machine shop did not get the cam nose back to spec because I had to shim the impeller out a little bit to keep it from contacting the cover.

 

[EddiePetty]

Re: Mercruiser 470 Water Pump Seal Leak

" I had to shim the impeller out a little bit to keep it from contacting the cover."

.......now that don't make sense to me. If the impeller were shimmed OUT....wouldn't that put the impeller closer to the cover?

"Why would you suspect oil seal trouble? I am running antifreeze at a 50/50 mix."

....in a previous posting you were concerned with heat build up....I was thinking along the lines of lack of lubrication: antifreeze is slicker than plain water.

Did you get around to taking run-out readings?

 

[Grub54891]

Re: Mercruiser 470 Water Pump Seal Leak

Bit of a reach but,is there sealer under the speedy sleeve? They do fit tight,and shouldnt leak,but it's a slight possibility.
Grub

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

Thanks everyone for all the great leads for solving this one.

Eddie, I meant the impeller was contacting the front cover of the engine, not the water pump cover. So it was as if the nose of the cam was too short, making it appear that the sleeve was not close enough to the end of the cam nose.

The first time I read your post I thought you were referring to the oil seal for the crank, thats why I was confused. I am following you now and yeh I know antifreeze is slicker than plain water.

I have to go buy a dial indicator and we had guest last weekend so still getting caught up with other things. Now getting ready for the 4th so I don't know how fast I am going to get to that run-out reading.

Grub, I did use locktite under the speedi sleeve. You are right, it is a tight fit and the cam nose was repaired so it has a good clean surface. I don't think I would get a gallon per hour through there even without the locktite. I am probably going to pull the sleeves off so that will rule that one out for certain.

 

[stonyloam]

Re: Mercruiser 470 Water Pump Seal Leak

Something you might do is go to Autozone (or some other parts store) and borrow their cooling system pressure tester. It fits in place of the pressure cap and you can pressurize the system to 15 lb or so and see if you can get it to leak without the engine running. At a gallon per hr it should be a substantial leak. Just a thought.

 

[EddiePetty]

Re: Mercruiser 470 Water Pump Seal Leak

.......still thinking about "shimming the pump impeller".

Any chance you installed a worn cam thrust plate when you reassembled the engine (or left it out completely!)

While you have begged/borrowed/rented/stolen a dial indicator to measure run-out....additionally measure the total float of the camshaft. A worn thrust plate may be allowing the cam to "walk" sufficiently to overstep the sleeves.

(Just thinking with my finger tips!)

 

[Argh!]

Re: Mercruiser 470 Water Pump Seal Leak

Terry, it is apart right now and after "fixed" it does not seem to leak until I get about 4 or 5 hours of run time on it.

Eddie, the cam thrust plate is installed and did not appear worn when I went through the motor the first time. It only had about 700 hours on it before the rebuild so all the parts were in good shape except the number 4 piston/rings that did not like the sea water from the rotted out exhaust riser elbow. Also, if the plate was worn then the cam would be able to walk toward the front of the motor making the cam nose appear too long, not too short. I pulled the sleeve off so there is no possibility of overstepping the sleeve when I put it back together.

I measured the cam run-out today. More than I expected, about 0.003" to 0.004" (min to max) but I suspect that I would get better readings if I pulled the cam out and measured it on a bench. Measuring the installed cam while cranking the engine I have various forces from lifting the valves. In hind sight I should have loosened all the rockers. In any case I don't think that is enough run-out to cause the leak even if it showed that much after a proper test.

The new front engine cover is on its way expected July 12, just in time for the weekend.