Definative main bearing thread

[bushmechanic]

Johns thread is full of good info but i noticed the last few pages have some people wondering what to do.

MAIN BEARINGS FAILURE, MUST CHANGE

The original bearings that are fitted to your to the range of Husaberg motorcycles are failing regularly the NJ206 bearing inner race will shed metal and put it through your engine and ruin the chrome bore and piston, KTM has the same problem on their EXC 450 -525 and use the same bearing...

husaberg.org

the ideas here are not just my opinion, there are a great many good minds who have worked out what causes the failures and even better have tried a number of different solutions. the majority of these solutions are scattered through KTM talks RFS section.

what causes the oem roller bearings to fail ?

1) crank spread under its own weight is the main problem

2) misalignment of the bearing bore, seems rare, search lineaweavers posts

3) insufficient crank axial play (end float) with crank spread

what is crank spread?

when the crank spins, its own weight causes it to spread outward opposite the pin
when you add in the combustion forces they contribute further to the spread.

sorry for the crappy pic there is a better one i will add later

in this condition while running the axial play of the crank reduces until the forces deforming the crank outward are matched by the crank resisting deformation (a function of its stiffness)

if you don't have enough end float or stiffness for the weight of the crank and the rpm its doing your end float (axial play) will become zero

with zero axial play the NJ series roller bearings will experience side loads as they are now subjected to some of the force causing crank spread

 

[bushmechanic]

NJ206 bearing design

first a pic of the failure

you can see there was about 1.5mm of the inner race unused and that the rollers were butted hard up against the lip of the inner race

schematic of the NJ bearing .. see the lip

an NTN bearing beside an SKF bearing note the SKF bearing has failed and the ends of the rollers on the SKF bearing have brocken off to the extent it almost looks like the same radius as the ntn

an SKF bearing note the sharp corners of the rollers compared to an NTN bearing

every single failure pic of these bearings shows the rollers were hard up against the inner race lip.

the NTN bearing has both nicer end radii and the lip itself has a bigger relief so that the contact with the rollers is closer to the centre

 

[bushmechanic]

axial play (end float)

there are 2 bearings one each side so when you measure the axial play of the crank you are measuring the distance between the rollers and the lip.

(you have 2 NJ series lipped locating bearing inner races one either side of the crank)

I ran 1mm endfloat

this is my ignition side Nj206 (ntn) after about 30 hours in the 700 build

 

[bushmechanic]

side loads

when the crank spreads and rams the rollers into lip on the bearing inner race they are subjected to side loading

this is OK up to a point but if its too much they will stop rotating and start to skid on the inner race ......

 

[bushmechanic]

solutions

what do you do about it?

1) first thing you can do is make sure the endfloat is big enough opinions differ but I suggest 0.5 - 0.6mm minimum for all the stock engines but i used 1mm on my 700

2) after a failure or during a rebuild check the crank isn't spread too much; correct the alignment or add pinch (opposite of spread) and..

3) weld the pin.. viewtopic.php?t=12083&f=5

4) use NTN mainbearings instead of skf ones

5) make the crank lighter reducing the outward loads due to the cranks own mass

6) use a completely different style of main bearing

 

[bushmechanic]

different setups

the first one that was popular was the use of a KTM crate engine deep groove ball bearing on the drive side and an NJ 206 on the ignition side.

many jap bikes use this idea but they fasten the ball bearing to the cases and also to the crank so the crank cannot move.

they also have about 2mm room between the rollers on the roller bearing and the lip or.. no lip at all so the rollers never contact the lip in these bikes

in the husaberg the ball bearing is just a stupdly tight fit in the cases and on the crank but the crank is not actually locked in position and can move.. you wouldn't know about it though till the roller bearing failed.

it is also very difficult to measure the space between the ends of the rollers and the nj roller bearings lip... you still need the same amount of room for the crank to expand because the axial loading is shared by the 2 bearings and the NJ bearing still gets squashed into the lip

it has worked for some people but i have seen it fail in 20 hours

 

[bushmechanic]

bigger drive side main

weed removed the counterbalancer and ran a 20mm wide nj2206 c4 "rollway" roller on the drive side with an nj206 on the ignition side

i tried it also with good results, need to make a spacer to replace the counterbalancer

 

[bushmechanic]

Other setups ACs

for a long time the accepted solution in the rfs engine was to use 2 Koyo or Nachi 6206 C3 ball bearings with a slip fit on the crank, not many people had any issues with these setups up to 570cc

there is also the possibility of using preloaded angular contact ball bearings like ducatis

as far as i know this has not been tried in a husaberg...

the issue is that the side loads normally contributing to crank spread are absorbed by the cases and IMHO the LHS case is not strong enough.

if you made a steel sleeve to better distribute the side load and reduced the weight of the crank to reduce the side load itself it would be worth trying simply because it prevents crank spread occurring at all. the code for the ACs is 7206B in abec7 watch the contact angle.. they are made in 2 different angles

some people have had good success with them in the rfs engines, others have broken their cases

 

[bushmechanic]

the ultimate setup IMHO is that tried by JBS and is a set of 22206 spherical rollers with pressure oil feed mounted in steel sleeves in the cases.

spherical rollers are cool because they can do this

I have one on the drive side with an nj206 on the ignition side and 0.8mm axial play..

 

[bushmechanic]

why the 550 ?

the majority of failures in Australia were the 550 engine

the crank is heavier than the 628 crank and also has a smaller pin, the engine revs more as well

so compared to the 628 you have more forces spreading the crank, less stiffness to limit the spread, typically less endfloat and SKF rollers with sharp radii

Orangeberg worked out the NTNs work better and stands by them as a fix.

this is also the consensus among the RFS engine builders, in fact the NTN bearings were used in the RFS engines as an OEM part for some years so they get to see the difference on a larger scale

 

[bushmechanic]

bearing codes

there is a lot of discussion in the main thread about bearing codes heat treatment and special steels

what it boils down to is that the only skf roller bearings you should consider using are the ones used as OEM parts, a standard SKF nj206 c3 or c4 is not good enough (aparently)

IMHO though you should not use SKF NJ seires bearings at all.. just use NTN or nachi or rollway.. they all have nicer roller end radii. the advantage of the skf (if any) is that they have 58 micron clearance vs 46 for the NTNs.

FWIW the ktm crate engine ball bearing is also a "special" TMB (transmission deep groove) bearing although many ktm engine builders have had great success with nachi and koyo C3 ball bearings

 

[bushmechanic]

so its simple when you understand the problem really comes from the crank not so much the bearings

As far as I know adequate endfloat + NTN rollers have been good enough for everyone worldwide who has tried them in a stock build, the same is not true for the SKF bearings. this is even the OEM skf bearings with the code NJ206ECP/C3HVC058.. they are not as good as the NTN used as OEM bearings in some years of the RFS with code NJ206 ET2XCS46

beyond that there are many other things you can do to improve on the standard arrangement

 

[Kye]

Bushmechanic, you're a legend mate!

I think you just described why I have hundreds of chrome looking metallic flakes on my strainer the last couple of of oil changes... and what to do to fix it. Greatly appreciated!

 

[zaga]

Hi Bushie

Great tread, with a very good documentation, thanks

In my old 400-02, I've been using 2 ball bearings SKF ETN9 C4, without a problem, what do you think about that solution ?

Also, what do you think should be my axial play ?

Thanks
:cheers:
ZAGA

 

[bushmechanic]

Hi Bushie

Great tread, with a very good documentation, thanks

In my old 400-02, I've been using 2 ball bearings SKF ETN9 C4, without a problem, what do you think about that solution ?

Also, what do you think should be my axial play ?

Thanks
:cheers:
ZAGA

I think if it works for you then it is very good, you are smart enough to have it setup right IMHO :cheers:

the axial play if its working.. then its good no? .. RFS builders use 0.2-0.3mm on that setup but they use a slip fit between the crank and the ID of the bearings, either by honing out the bearing ID or polishing the crank down. this is done so the crank can't pull the bearings out of their bore and wear out the cases and also so that if the bearings move on the crank when they are hot it cannot be locked without enough end float when it cools back down.. it greatly reduces the axial loading on the bearings.

with the C4 clearance this could be less of a problem but is still something to consider.

check how the bearings can be pulled inward out of the cases 8O

Interesting experiment i tried to retain the outer race of an Nj2206 with an alloy plate

you can see its cracked. the outer race of the bearing has been pulled
out of its bore by 1.5mm and with enough force to break the retainer.

I find this interesting as there is no easy way to apply a force to the outer race
in this direction. there is a lip but it can only push the bearing into its bore not
out

also I was running 0.8mm crank endfloat so for the outer race to move any
more than 0.8mm out of its bore means there is some extraordinary flexing
going on.

im guessing that at TDC on the exhaust stroke the piston and rod pull hard
enough on the crank to make it flex inward and at the same time apply enough
radial force to the rollers to let them lock up and pull on the outer race.

in any case because of this experiment and so I can measure the axial play
properly Ive honed out the ID of the spherical roller about 2 thou in total
to be a 1/2 thou slip fit on the crank.

 

[zaga]

Hi

My endfloat is 0,3 so it should be in range, a friend mechanic told me it could be even more without a problem.

Thanks
:cheers:
ZAGA

 

[bushmechanic]

yes your right I reckon 1mm is pretty safe

regardless of the type of bearing used the crank still needs the same amount of room to expand

well.... the ACs and the spherical rollers don't need any axial play but the cases do

 

[bushmechanic]

Bushmechanic, you're a legend mate!

I think you just described why I have hundreds of chrome looking metallic flakes on my strainer the last couple of of oil changes... and what to do to fix it. Greatly appreciated!

Kye this a drain plug after 1 main failed in a 550

if the mains are ok just a few flakes can also be the bigend

 

[contiman]

JBS kit on my 710 husaberg,
I run 22206 crank bearings, welded pin crank
2005 special crank

welding of the pin is cracked after 25 hours!!!!

 

[Kye]

Kye this a drain plug after 1 main failed in a 550

Whoa! Yep, well I reckon that is what my drain plug would look like if I had a magnetic one (must throw one of those on the shopping list).

I had no idea these cranks flexed so much!