Crankshaft Misalignment Repaired

[Captain Jerry]

The biggest diesel engines made do not have one piece crankshafts.  They are built up of forged webs and journals assembled by shrink fit of the webs onto the journals without the use of pins or key ways, much like many of the small engines built on this forum.  What happens when one of these built-up crankshafts slips around the journal?  In the home shop, you might put it in the vise and straighten it out or make a new one.  What do you do if the crankshaft is in a marine diesel, in a freighter half-way around the world?

READ THIS!  It is an well told story of heroic engineering that should be read by anyone who has ever looked at a damage engine and said "Oh sh&t!

http://users.telenet.be/doxford-matters/tekst/Re-alignment%20Eastern%20Rover.pdf

[dsquire]

Captain Jerry

Gerry Thanks for posting that. It was a good read. It seems that no matter what needs fixing that is impossible next to impossible to fix there is always someone that can figure out a way to get the job done.

Cheers 

Don

[smfr]

LOL, they had me at "Using four tonnes of liquid nitrogen"

Simon

[steamer]

Imagine the "fortitude" required to propose that solution!....Very cool!

Dave

[steamer]

Here's what the Doxford cranks looked like....

http://users.telenet.be/doxford-matters/crankshaftbay.html


Dave

[steamer]

An interesting read Jerry...

http://users.telenet.be/doxford-matters/tekst/Barclay%20Curle-Doxford.pdf

Dave

[ths]

That book must have been an international best seller. Da Vinci Code? Hah!

Hugh.

[Captain Jerry]

Dave

Thanks for the links.  I have been dropping in on the works link for several months or maybe years.  I never get tired of looking and almost always get some new piece of information from it.  I have not seen the operations manual before and will take some time to read it through.   

I have been thinking through the repair job and the  information is very detailed but there is one big question that I could not get resolved until I took one more look at the works pics in your link.  I have the answer now but first I'll ask the question and see anyone else knows the answer. It may be obvious and I'm just thick headed but the clues are there in the works link.

Here is the situation.

There are three crank throws for each cylinder.  The lower piston drives the center crank and the upper piston drives both of the outside crank pins through side rods.  The center crank is offset from the two side cranks by 180°.  The dogbone shaped main webs hold the center crank pin in the bore at one end and holds the side crank in the bore at the other end.  The misalignment occurred in joint between the big dogbone web and the forward crank pin for the side rod of cylinder four   All of this is stated in the report,  I'm repeating it here so you don't have to go search for it in the report.

The fix was  to shrink the diameter of the side rod crank pin so that it would be loose enough in the bore to be jack back into alignment by applying force at the opposite end.  A bridge/brace was placed under the side rod crank of cylinder 3 to prevent rotation and a vertical jack was placed on top of the main shaft journal between cylinders 3 and 4.

Here is the question.

The nitrogen chilled the entire pin.  What prevented the other end of the pin from rotating in the bore of the half web when the jacking force was applied.   Here is a hint.  They weren't worried about it. They never even checked the witness mark at the other end of the pin.

Why not.

I have an answer that I am pretty confident of that I will post in a few days.  I could be wrong. It has happened before.  Your answer could be better.

Jerry                                                                                                                                                                    224

 

[Captain Jerry]

This is what a two cylinder Doxford crankshaft looks like.  You can rotate the 3D PDF by clicking on the activate button at the top left of the picture.   This is not to scale but to illustrate the arrangement of the crank and main journals.

[Dan Rowe]

Here is the question.

The nitrogen chilled the entire pin.  What prevented the other end of the pin from rotating in the bore of the half web when the jacking force was applied.   Here is a hint.  They weren't worried about it. They never even checked the witness mark at the other end of the pin.

Jerry,
The last sentence of page 14.

"An initial calculation showed that a temperature difference of approximately
150ēC would produce relaxation of the shrink and this could be achieved by
applying gentle heat to the web in the final cooling stages of the crankpin."

They only applied the heat to the crank web at the forward side of the pin so there was little danger of the aft end slipping as there was not a 150ēC temperature difference at that end. Just because there is no mention of the aft witness mark is no reason to assume it was not checked.

Dan

[steamer]

That's what I was thinking Dan.....
Getting something that cold is actually pretty hard to do....
Dave

[Captain Jerry]

They only applied the heat to the crank web at the forward side of the pin so there was little danger of the aft end slipping as there was not a 150ēC temperature difference at that end. Just because there is no mention of the aft witness mark is no reason to assume it was not checked.

Dan

That is what I first thought, Dan, but then I looked a little closer.  The thickness of the side rod web, (side crank pin to main shaft) is much less than the thickness of the big dogbone web.  It should then be correspondingly weaker.  Why did it not fail before the crank pin/dogbone joint? And the pin was chilled with only a slight differential made up by heat to the larger dogbone.  If there was any chance of it shifting, I think it would have be prepared for, just as they prepared for over correction by having a jack on the opposite side to bring it back. 

I am glad that you looked at it, Dan.  There are only a handfull of people on this forum with your experience in this area.  I'm not saying that you're wrong but I think there is a better reason.  I am still searching through the available information for more support of my theory.  I'll post it all shortly.

That's what I was thinking Dan.....
Getting something that cold is actually pretty hard to do....
Dave

Come on, Dave. Try a little harder.  Four tons of liquid nitrogen!  Even a brass monkey would lose his grip and his gonads.
I'll leave the question open for another 24 hours so other members have a chance to weigh in.   I will be surprised if no one else comes up with another opinion.

Jerry

[steamer]

Jerry....How much does that crank weigh?    Thermal mass is not held in church!

You will be sucking the heat out of everything it touches....and it touches a lot.

....but I'll look...

Dave

[Captain Jerry]

Dave

I guarantee that Arnold 'Daly' Duncan know the answer.

Jerry

[steamer]

If it was immersed in 4the tons of LN2.....id be with you Jerry
But it all pissed thru a little hole inside of a large piece of shaft which was being heated by a pair of 1" rosebuds....now at ive heated 9" steel shaft with 2 of them.....and if I had
4 tons of Ln2 going thru it...id say it would be a battle.

Whats missing from this eqaution is the resistance to heat transfer across the interface between the two parts....which is not insignificant.   Thats what made it work.

Dave

[Maryak]

As I see it, everything to the right of the pin is locked, everything to the left is free so when the jacking force is applied, its applied to the "free" side this I think leaves only the web under question to move relative to its loosened pin. Hence the bottom of the web is most unlikely to have moved relative to the main shaft also there is no rotational force being applied to the pin, only the web and everything to the left of it. Non the less I would have wanted witness marks on every web/pin interface of the No4 Unit.

Best Regards
Bob

[steamer]

Ditto!

Dave

[Captain Jerry]

Dave

There is a lot of uncertainty in the process.   First, the target temperature is 70° C below the range of the alcohol thermometer used to measure with so the temp at the interface was interpolated from the slope of the charted temp decline. The pin was pre-cooled with dry ice for 30 hours before the nitrogen flow was started.  The nitrogen was flowed through the pin for nine hours. The "gentle" heat was applied to the web for only about 15 minutes.  This was more than two hours after the last actual temp was known but there was now way to know what actual effect the heat had on the temp difference  between the pin and the big web since the heat reversed the slope of the temp decline used to estimate.

It worked.  The joint was relaxed and the web was moved.  But there was no way to what the temperature differential was at the other end of the pin and whether that joint had also relaxed and if both ends of the pin had relaxed, was it not possible that both joints would slip to some degree under the force of the jack.  If the thickness of the short web at the opposite end of the pin is smaller than the thickness of the big web, the contact area would also be less and therefore that end pin might slip under less force.   Certainly, the cooling worked and the repair was a success.  The question that I had was why was there so little concern that the opposite end might slip. 

As Bob says, I would want to check witness marks all over the place checked.

Dan says that there may have been concerns and measures taken that were not mentioned in the report but the report was very specific about the steps taken to stabilize the crank shaft, including a third jack applied to the main journal between #4 and #3 and the bracket girder under the side rod pin of #3 which is three shrink interfaces away from the repair.  I think the report would have mentioned the joint at the other end of the pin if there was any possibility of it moving.

Bob

The pressure on the jack was raised to 200Kg/cm with no result.  At least that much torque was transmitted to the pin.  The joint did not begin to move until the pressure reached 499Kg/cm so quit a bit of torque was applied to the pin and carried all the way to the brace under the aft side pin of #3.   I think there could have been considerable risk of movement at any of the intervening shrink joints.

Now here is a partial answer.  The report misstates the situation. As I said before, Arnold 'Daly" Duncan know the answer.  It is right in front of him.  Since raising this question, I have researched the problem and I find many pieces of evidence to support my thinking. 

Here is another hint.  In all of the pictures in the crankshaft shop,  how many pictures of individual dogbone main webs do you see in various stages of fabrication and how many pictures of individual side rod webs do you see.

I'll quit jerking you around tomorrow.

Jerry

[Maryak]

how many pictures of individual side rod webs do you see.

Jerry

None, because I think they are integral with their associated main journal,

Bob

[steamer]

"It worked.  The joint was relaxed and the web was moved.  But there was no way to what the temperature differential was at the other end of the pin and whether that joint had also relaxed and if both ends of the pin had relaxed, was it not possible that both joints would slip to some degree under the force of the jack.  If the thickness of the short web at the opposite end of the pin is smaller than the thickness of the big web, the contact area would also be less and therefore that end pin might slip under less force.   Certainly, the cooling worked and the repair was a success.  The question that I had was why was there so little concern that the opposite end might slip."

Good point Jerry....the engineer wasn't heating the other web...just the one in question...which is where the pin with the LN2 was ....and no there was no garrantee the other web WOULDN"T turn...if you look at other pages in the site, there are some "BLOWN ENGINE" pictures that are pretty spectacular...look at what they had to do to change the parts out...basically cut a big hole in the decks.  A GREAT DEAL of money in a foriegn port....if this worked...they fix the crank in place, and "off you go"....if not....get the torch...IMHO...it was a calculated risk that the crank web they focused on would be the only one to move.   Physics is working in your favor, and you can justify the cost of ripping the deck off if and only if you PROVE that the crank couldn't be repaired in situ.   My guess is the pucker factor was mighty high when the jacks were being pumped up!  There was a lot riding on it.

Dave


 

[Dan Rowe]

Jerry,
Take a look at this 4 cylinder Doxford crank.
http://www.oldengine.org/members/diesel/marine/doxford2.htm

Here is a verbal account with some photos of doing a similar repair.
http://www.cultureshock.org.uk/stories/crankshaft-repair-in-shanghai.html

I agree with Dave they made the repairs in 7 days. If it did not work the next step is to take a gas ax and remove the house to get room to remove the crank.

I was on a ship where the 3 diesel generators were replaced while underway. We had containers with generators on deck while the work was carried out. It was lucky that there was a hatch to the cargo space directly above the generator room. They had a big shore crane to lift out the old engines. The work was all done with minimum disruption of the cargo operations.

Dan

[Dan Rowe]

As Bob says, I would want to check witness marks all over the place checked.

Jerry,
The witness mark to tell them that the crank had slipped 9.5mm was made at the factory and each shrink fit on the crank has a similar witness mark. I fairly certain that it is standard practice to check all the marks after an event of this magnitude.

Dan

[Captain Jerry]

how many pictures of individual side rod webs do you see.

Jerry

None, because I think they are integral with their associated main journal,

Bob

  That's what I think too.  It is a single piece forging from the pin in #4 to the pin in #3.   The other end of the pin could not rotate because it is not pressed into the side web.  It is part of the side web.  Did I mislead you into thinking that it was a separate pin?  Well, maybe, but the report and many other descriptions of the crank constrictions describe it as fully assembled with pins pressed into webs.  That was what lead me to start thinking about the details of the repair.  If, like the engineer in the report, I had known that the side crank pins were not a shrink fit in the side web, I wouldn't have worried about it either. 

The earliest picture that I can find of a Doxford crank seem to confirm that. There is a picture of this in the link provided above by Dan. The side webs have much more metal surrounding the pin and shows the main journals with flanges that bolt to the web.  The bolt on journals made a much longer shaft and the early engines show much wider spacing between cylinders.  In these engines the upper and lower pistons had the same stroke and the side throws were equal to the center throw.  With the introduction of the balanced "B" and "LB" series, the throw of the side rods was reduced to give better balance and reduce vibration.  If you assume that the side web was a separate piece with two holes in it for the crank pin and the main journal, which is the way that you and I would go about building it, when the throw was reduced, these holes come much closer together and may even overlap.  That would seem to make it impossible to shrink fit to two pins.

The forging was not a Doxford process.  The forgings were bought in.  There are no pictures of this process in the Doxford works because it wasn't done there.  The closest that I can find is the picture in which Arnold 'Daly' Duncan shown machining the forging shown in the group of pictures of the Doxford Crankshaft Shop.  You can see the crank pin to the left of his head, the side web already faced, and the main journal being machined.  Both pins are integral with the web.

As further evidence look at the drawing at the top of page 17 and you can see that the side crank pin (the one in question) and the #4 journal almost overlap.  They were not likely to be separate pieces in a shrink fit.

I know this is a lot of conjecture but  I was just showing the amount of hidden detail contained in the Doxford Works photos and why I enjoy revisiting it.  The first time that I looked at the photo of Alan Duncan machining the forging, I was concentrating on the size of the equipment, the details of the tool holder, the details of the carriage and cross feed, and the operators position.  I must have looked at it a dozen times before I began to focus on the part, and it was not until I read the repair report that I really understood the part.

I think I have discovered a link to a Doxford design engineer and I am attempting to get confirmation of the configuration of the forging.  The link was valid 4 months ago so maybe I will get a reply.  I'll post it if I do.

Jerry

Jerry

PS.  If anyone is tempted to say "Of course. That was obvious" let me remind you that I said the answer may be obvious to some but until now, no one has said so.

[Dan Rowe]

Jerry,
I see your point....and yes I missed it. The location of the plug for the nitrogen and the pin are a single piece. The girder is the rotation stopper on the same forging and the jack with a copper backer in the bearing is simply insurance.

Here is a better link to the works photos:
http://www.shipsnostalgia.com/guides/William_Doxford_and_Sons#The_Manufacturing_Process

The forging in question can be seen in one of the overview shots and in the lathe as you mentioned.

Thanks for bringing this up. Very interesting.

Dan

[ScroungerLee]

No one has mentioned that the joint that moved was obviously the weakest link in that part of the crank, so why would some other place move first?  I do agree with the pin being part of the side web.

Just my thoughts, I have little knowledge to go on.

Lee

[steamer]

OK Jerry,,,,I think I understand what you getting on about.    The crank stopper on the other side of the journal bearing isn't in torsion during the "push", but in bending.....for those of you following along...look at the top of fig 6A in the story on page 1 of this thread.


Under those conditions, its impossible for the pin joint there to rotate about the axis of the pin.....I would want to be sure it wasn't bent though.


Is that what your getting at?

Dave

[Dan Rowe]

Dave,
It was Fig 6A I was going on about. The section of the crank from the nitrogen outlet to stopper E is a single forging.

They did check for bending with the crankshaft deflections.

Dan

[steamer]

Yes   I agree about the forging...the orientation of the next joint where the stopper girder is,  is what I missed.
Still took some big cohones to pull that corrective action off....but there was nothing to loose....so what the hell.

Dave

[Captain Jerry]

Still took some big cohones to pull that corrective action off....but there was nothing to loose....so what the hell.

Dave

Maybe not so big.  Dan turned up another report of a similar repair on a Doxford crankshaft.  It was done without liquid nitrogen, only dry ice and oxy-propane torches.  The misalignment may not have been as great, since the ship had completed a voyage with the crankshaft in that condition.

Still, the Eastern Rover was the first such repair attempted, so probably big cohones.  After that, it seems to have been done one or more times, allowing smaller cohones each time.

Jerry

[Dan Rowe]

Jerry,
You might not know of the locomotive steam cranes used at the Doxford works. Here is a link with some photos and specs.

http://homepage.ntlworld.com/foxfield/roker.htm

Dan

[Captain Jerry]

OK Jerry,,,,I think I understand what you getting on about.    The crank stopper on the other side of the journal bearing isn't in torsion during the "push", but in bending.....for those of you following along...look at the top of fig 6A in the story on page 1 of this thread.


Under those conditions, its impossible for the pin joint there to rotate about the axis of the pin.....I would want to be sure it wasn't bent though.


Is that what your getting at?

Dave

Not quite , Dave.  I was certainly not concerned with rotating the side crank pin on #3.  Did you know before that it was a single piece forging that included the side webs and pins of both #3 and #4 as well as the main bearing journal between them?  I didn't and all previous descriptions of the crankshaft said that it was completely assembled which lead me and maybe some others to think that both ends of all crank pins were a shrink fit.   They were not and that is what I discovered for myself and maybe some others from doping out the hidden clues contained in the Doxford Works photographs.

Jerry

[Maryak]

all previous descriptions of the crankshaft said that it was completely assembled.
Jerry

I suspect this was completely true in the original design where both pistons had an equal stroke.

Best Regards
Bob

[Anzaniste]

Here is a better link to the works photos:
http://www.shipsnostalgia.com/guides/William_Doxford_and_Sons#The_Manufacturing_Process

Dan

What a fascinating series of photographs. It really brings home how much knowledge and experience has been lost.

[IanL]

Regarding the building of the crankshaft, The earlier type was simplified in the mid 1960's whereby the main journal bearing and the adjacent webs was incorporated into a single large diameter bearing.  In the case of the 760mm bore "J" type engines the main journal was 1150mm in diameter and about 300mm wide.  It was machined by feeding in a 300mm wide tool on the lathe.  A tightly wrapped roll of steel foil was the result. Perhaps the widest parting off tool in the world!

So this section of the crankshaft was effectively the journal described above with a pin on each side of diameter 625mm sticking out by about 600mm.

The centre part of the crankshaft then designated the crankshaft as either "Semi-built" or "Fully-built".  The fully built was then formed from two huge thick plates with a 625mm hole at one end and a 640 diameter hole at the other end.  These were temporarily joined together for machining on a large vertical borer.  Then they were separated and the 640mm bore end heated with gas burners.  (Always turn off the gas supply before the air supply or just lift the burner out whilst burning correctly, otherwise the carburising flame deposited carbon on the inside of the bore which was tricky to remove and reduced the friction between the web and the pin which could lead to a slipped shrink in service).  Each heating sequence took about 6 hours to heat and a few minutes to insert the pin followed by a long cooling period so about one shrink per day was possible.  Once the two webs were together with the centre pin, then the whole crankshaft was shrunk together in modules.

The semi-built version had the centre webs and pin provided as a single forging, but the rest of the assembly was the same.  Machining the centre pin on the vertical borer was quite tricky as the tool had to reach along way between the webs.  Semi-built webs were mainly used on the smaller bore engines, i.e. 580mm and 670mm and these gave smaller polar inertia values.  I think that the 670mm bore engines could use either type and this allowed selection of torsional vibration natural frequencies to suit the application.

I hope that this explanation has helped, I was an apprentice at the time and it seems so long ago now (1973)
Ian