Sealing crankcase halves

[dxhall]

I’m about ready to do the final assembly on the Harley Sprint / Aermacchi 350cc single cylinder motor.  The gasket set includes a paper gasket that goes between the crankcase halves, but the Yamabond sealant I normally use is intended for metal-to-metal contact, not to supplement paper gaskets. Would you use the sealant only, or would you use the paper with sealant too?  I suppose I could use a conventional sealant, but the Yamabond by itself comes highly recommended.

[Scout63]

I’ve never built that engine, but I swear by Yamabond for Norton cases and Honda single cam covers.  Always without gaskets and no problems.

[lucky phil]

I’m about ready to do the final assembly on the Harley Sprint / Aermacchi 350cc single cylinder motor.  The gasket set includes a paper gasket that goes between the crankcase halves, but the Yamabond sealant I normally use is intended for metal-to-metal contact, not to supplement paper gaskets. Would you use the sealant only, or would you use the paper with sealant too?  I suppose I could use a conventional sealant, but the Yamabond by itself comes highly recommended.

AS long as removal of the paper gasket has no clearance effects then just use the Yamabond. Even if it does and you can re shim for clearances thats acceptable.

Ciao

[mechanicsavant]

+1 on yamabond , there are a few different strengths too .

[Dukedesmo]

I like threebond, never had a leak and it seals even where the surfaces are not perfect but, not fitting a gasket between the halves will mean the crank etc. is tighter so you 'll need to shim it accordingly.


That said rebuilding even with a gasket would still require endfloats/preloads to be checked and altered if necessary so no more work really.

[Moparnut72]

When I worked in the aircraft shop it cost the owner extra to get his engine overhauled ifit was assembled with Yamabond. It was extremely difficult to separate cases put together with that crap. It took extra time to get cases apart to say nothing of trying to clean them of that stuff. Use the paper gasket, apply Ultra Copper to both sides smearing it on lightly with your finger. Won't leak and will come apart easily if you have to go back in. We never had a leak on any of our radials. The only leak we would ever have would be from a valve cover that wasn't resurfaced properly.
kk

[lucky phil]

When I worked in the aircraft shop it cost the owner extra to get his engine overhauled ifit was assembled with Yamabond. It was extremely difficult to separate cases put together with that crap. It took extra time to get cases apart to say nothing of trying to clean them of that stuff. Use the paper gasket, apply Ultra Copper to both sides smearing it on lightly with your finger. Won't leak and will come apart easily if you have to go back in. We never had a leak on any of our radials. The only leak we would ever have would be from a valve cover that wasn't resurfaced properly.
kk

As an aircraft engineer with experience working on Radials and flat engines as well as World Superbike engines and engines that have competed at events like the Isle of man I've only got 1 thing to say. He's not working on an old 1830 or R2000 engine. Nobody uses old composite gaskets anymore on modern motorcycle and automotive engines and they replace them with modern sealant anytime they practically can if there're smart. If they do use gaskets for access covers etc it's embossed metal gaskets. As much as I love old a/c engines it's ancient technology with ancient practices. Engineering moves on as does materials.

Ciao

[LesP]

I’m about ready to do the final assembly on the Harley Sprint / Aermacchi 350cc single cylinder motor.  The gasket set includes a paper gasket that goes between the crankcase halves, but the Yamabond sealant I normally use is intended for metal-to-metal contact, not to supplement paper gaskets. Would you use the sealant only, or would you use the paper with sealant too?  I suppose I could use a conventional sealant, but the Yamabond by itself comes highly recommended.

Its an ancient thing with a wide band mating joint so unlikely to leak unless the surface is damaged.

If the gasket is from quality paper why not use it.
All I would add is and only my method of that job (with or without a gasket)

Any riser on the mating surface needs to be addressed and can be peened back (over removal leaving a gouge)

The surface can then be lightly surfaced (I use a 16 mm square lathe tip tool holder with medium grade W&D dry only in a direction of rotation, not across the mating face.
Any gouge / nicks (as an option)  can then be filled with Devcon or similar (JBW) and resurfaced.

Any fastener hole and thread entry at the joint can then get a light countersink so the fastener is only applying clamping force with no impact on the joint face. (I do the same if it is studs)

The faces can then be cleaned with Acetone (time after depends if filler has been used)

Paper gaskets get a light wipe with Loctite 515 or 518 (They are basically the same)
If you want the joint to last forever, Wellseal can be used instead, that would be a light application to all four surfaces (case and gasket) left for 25 minutes to tack off then the joint assembled and torqued to spec.
That would last longer than nuclear fallout.



I also use Kevlar thread on some joints on occasion (over wide dental floss in the past)

[Moparnut72]

Its an ancient thing with a wide band mating joint so unlikely to leak unless the surface is damaged.

If the gasket is from quality paper why not use it.
All I would add is and only my method of that job (with or without a gasket)

Any riser on the mating surface needs to be addressed and can be peened back (over removal leaving a gouge)

The surface can then be lightly surfaced (I use a 16 mm square lathe tip tool holder with medium grade W&D dry only in a direction of rotation, not across the mating face.
Any gouge / nicks (as an option)  can then be filled with Devcon or similar (JBW) and resurfaced.

Any fastener hole and thread entry at the joint can then get a light countersink so the fastener is only applying clamping force with no impact on the joint face. (I do the same if it is studs)

The faces can then be cleaned with Acetone (time after depends if filler has been used)

Paper gaskets get a light wipe with Loctite 515 or 518 (They are basically the same)
If you want the joint to last forever, Wellseal can be used instead, that would be a light application to all four surfaces (case and gasket) left for 25 minutes to tack off then the joint assembled and torqued to spec.
That would last longer than nuclear fallout.



I also use Kevlar thread on some joints on occasion (over wide dental floss in the past)

We restored our engines, they went back together  exactly the same as they came from the manufacturer. No, cases didn't get gaskets,  one make that I don't remember got silk thread. I was referring to those surfaces that did, valve covers etc. However Ultra Copper wasn't available back then but no engine left our shop with so much as an oil weep. The remark that radials mark their territory like old Harleys is not necessarily the case with a properly assembled radial. I wasn't an AP so I didn't do any final assembly, I disassembled engines, cleaned, painted, measured and prepped parts.
kk

[Chuck in Indiana]

one make that I don't remember got silk thread.

Continental..

[Moparnut72]

That sounds right. I didn't work in the clean room where the engines were assembled. I worked with the bead blaster, parts washers and all kinds of dirty stuff. Working there got me into wearing black clothing, nothing like dirty non-detergent oil.
kk

[dxhall]

Thanks for all of the advice.  There are really two issues: (1) best way to seal the cases, and (2) how to maintain correct crankshaft endplay.  The endplay issue has me stumped. 

The factory spec is .0005 to .0025mm.  The paper gasket is .2mm thick.

On a motor with shell-type bearings,  the crank can be moved forward and back with only hand pressure.  Measuring endplay only requires reading a dial indicator on the end of the crank.

On a motor like this, though, the crank runs in ball bearings, and the crank journals are a press fit in the inner race of the bearing.  So, since the bearing is fixed in the case, and the crank is fixed in the bearing, the crank doesn’t move forward and back.

Here’s a photo in the manual on how endplay is measured:







The factory tools must attach to the crank somehow and pull the crank forward and back.  I, however, have no such tools, and don’t see how to make them. 

Tapping on the end of the crank with a wood block / hammer setup wasn’t enough to force the crank into the bearing.  So tapping on the ends of the crank to measure how much it will move won’t work.  I had to use a press to get the crank into the bearing.

I don’t see a way to get the crank to move so that I can make this measurement.  So - does it really matter, so long as the cases will go together?

[lucky phil]

Thanks for all of the advice.  There are really two issues: (1) best way to seal the cases, and (2) how to maintain correct crankshaft endplay.  The endplay issue has me stumped. 

The factory spec is .0005 to .0025mm.  The paper gasket is .2mm thick.

On a motor with shell-type bearings,  the crank can be moved forward and back with only hand pressure.  Measuring endplay only requires reading a dial indicator on the end of the crank.

On a motor like this, though, the crank runs in ball bearings, and the crank journals are a press fit in the inner race of the bearing.  So, since the bearing is fixed in the case, and the crank is fixed in the bearing, the crank doesn’t move forward and back.

Here’s a photo in the manual on how endplay is measured:







The factory tools must attach to the crank somehow and pull the crank forward and back.  I, however, have no such tools, and don’t see how to make them. 

Tapping on the end of the crank with a wood block / hammer setup wasn’t enough to force the crank into the bearing.  So tapping on the ends of the crank to measure how much it will move won’t work.  I had to use a press to get the crank into the bearing.

I don’t see a way to get the crank to move so that I can make this measurement.  So - does it really matter, so long as the cases will go together?

Here's how to set crank end float on a ball race main bearing engine. This is how I do it on a Ducati Bevel drive and belt drive V twins. Install the crank without shims and bolt together the two case halves without a gasket because that's how I assemble them or with a gasket if you must. Turn the cases on their side with the crank free and set up a dial indicator on a magnetic base attached to any piece of flat scrap steel you have screwed to the outer engine cover mount holes. Lift the crank up and down and take a reading of the end float. Now a Ducati engine uses angular contact main bearings which are designed to take axial load as well as radial and Ducati specs recommend .15mm or .006" preload on the crank. So the cases need to stretch .006" on a cold engine. The reason for this is that when the engine gets to operating temp and the cases expand at a greater rate than the crank you will end up with close to zero bearing preload or around .001-.002 preload. You then seperate the cases and add the shims required and bolt it up again and check the crank drag. You can't now measure the crank float because there's preload but when you turn the crank if you have enough resistance for the crank to just barely hold itself with the bob weighs at 90 degrees your going to be right. You want some bearing drag which will disappear when the engine reaches operating temp by and large. Ducati always had them shimmed too tight from the factory for some reason. So when you're trying to achieve this type of precision you may be able to see why a centre gasket is a PITA to some extent. If you use one you then need to have another identical gasket on hand for the final assembly and the gasket itself is another variable you can do without when you can use a modern sealant that has no compressibility variation issues or requires any compensation at all for end float setting.
In the case of a plain ball bearing engine I would still set it up with a few thou interference because it will disappear when it's warm and a plain ball can accept a bit of axial load without issue.
Dial indicators and mag bases are cheap these days or you can borrow one. In some cases if you have an engineering flat and the instillation suits you can use that and feeler gauges but that's going to be pretty rare.
In your case there's a few questions. Is it the std practice to press the bearing onto the crank journal or this engine? Is this the manual procedure? If so then instead of the tool shown you may be able to just bolt some steel plates to the engine cover mount holes and use a couple of large screwdrivers or tyre irons and lever the crank to get a reading. Without knowing the details of this particular engine I'd say yes the end float is necessary to do but like most things it depends on the "degree" of deviation from the limits that's acceptable. Getting this stuff right is the difference between a well built engine and a backyard hackers rebuild.
The advantages of sealant are, no compressibility issues, simpler end float shimming, simpler assembly, in some cases simpler cleaning (ever had to scape off some old gaskets that were baked onto the cases) no fretting of the gasket under vibration (more a race engine thing) flexibility to adapt to hot cold cycling and consistency of quality. Sometimes you don't know what the quality of an aftermarket composite gasket really is, looks ok but could be any raw material in reality, better ability to seal imperfectly matched surfaces.

Ciao