Josephine rebuild

[Keith Clements]

The block and heads having passed the porosity and leak test so I moved on to the crank.
You cannot be certain of the state of any component so you have two choices 'suck it and see' or 'check it out'. The latter is almost impossible to do completely unless you have a battery of facilities available.
The crank... has it any cracks ? .. is it dimensionally correct? ...has it the correct hardness? ...is it balanced?
You can strike it and see if it rings correctly, you can do magnetic or chemical crack testing.
It may have been reground but was it done correctly, to the desired tolerance and without inducing stress points?

I tried a few ways of mounting the crank in the lathe. A live centre in the tailstock and a centre in the headstock. Then I made up a bearing to fit inside the rear housing where the clutch input shaft goes. Then a three jaw chuck on the shaft itself. All these were attempts to reduce run out. Currently it is about .o4 mm or 1.5 thou so I am trying to reduce this. I may have to go back to checking out the lathe bed and tailstock alignment as it may have moved since it was set up years ago.



And then a dynamic test.
TOPIC

There was no perceptible vibration at this speed but I would like to quantify any there is. My very sensitive DRO should pick up any so in the absence of strain gauges, velocity transducers and other sensors that make up a professional balancing system I will see where I get. I have been looking at various designs of equipment and seeing how they achieve balance. My worry is that all this very expensive equipment is only as good as its operator who may be trained to balance a Chevy V8 but a Jowett flat four?

[Srenner]

To avoid the lathe's out-of-center showing up in the crank, most shops use precision "Vee" blocks. Here's a video of straightening a crank, but at 40 sec, the instructor checks the crank for straight. This is usually done after grinding so that the journal is known to be round.https://www.youtube.com/watch?v=NSnzdWSgEFs

[Keith Clements]

Thanks Scott. Looks a bit stress inducing that method , in more ways than one.
Yesterday I used another method and put a centre and faceplate on the headstock.

Then I found that the centre on the flywheel was not good so put the oilite bush in and centred on that. Much better now less than a thou run out.

Then I used the V blocks to create a knife edge so I could check static balance. I will do some experimentation to see how accurate in grams I can balance.

Having now seen your video I will see if it is straight also, although as it runs well in the block it should be OK, although the V blocks I have are too wide to fit in the journal.

[Keith Clements]

Ordered my Plastigauge today £54 after VAT and postage ---a bit more than the last time I bought it.
http://plastigauge.co.uk/shop/plastigauge-pl-x/

This measures from just under .075 thou (which was the recommended Jowett setting to reduce crankshaft breakage) to 1.77 thou.
Scott and Neil had something to say on bearing clearances with Neil favouring building as tight as possible, Indeed most engines I have built have been difficult to turn initially due to overcoming the stiction. Then they require strong hands to keep them turning. Since it was twenty years since I last fitted bearings I am having to relearn the technique. The engine I did build that long ago still had 75 psi oil pressure and even above 50 psi when very hot. So when I get around to putting that back in another block I will check the clearance. For now I am checking out another crank for another block.

I have given up on trying to static balance the crank/flywheel assembly as even a bolt added to the outside of the flywheel cannot make the thing turn, so the chance of getting balanced to within a gram is unlikely! That was tried between centres and on knife edges and sat within a block in its bearings. Perhaps a Teflon sheet would reduce the friction or a couple of extremely hard knife edges would provide the desired low friction support.

So some more research on dynamic balancing will ensue.

[Srenner]

Wow! I buy just one of the little paper packages for about US$2.50!

[Keith Clements]

Having researched some more on engine balancing I am convinced the Jowett method of balancing the engine after the test run is the way to go. The reasoning goes like this. The flat four does not need complex external balancing like a V8 as it is inherently a balanced design. Also the guess made by most crankshaft balancers on the bob weight is just that. Many often add 'a bit' they call 'over balancing' because the initial guess at 50% reciprocating weight was not correct. Such a correction may not apply to a flat four. The other forces acting on the piston when it is in real operation such as ring scrape considerably alter the reciprocating force (dynamic mass) on the crank from that guessed at in setting the bob weight.

Sure having statically balanced con rods and pistons is required. The usual way of measuring the con rod little end and big end mass is to pivot on one and measure the weight on a scale at the other end. Of course the pivot needs to be close to frictionless and is usually a polished pin that fits the respective journal. The centre of gravity of the con rod needs to be found (by balancing on a string) and then the total weight proportioned accordingly. This proportion can be used to check the mass found in the two weighings done previously. Mass can then be removed from the respective ends of the heavier rods.

A separately balanced crank and flywheel also helps particularly if they have been machined. But as I am using previously used components the crank should be good but the flywheel and drive belt pulley may have been matched to another crank and thus be out of balance statically and dynamically with that crank. I have not looked for quotes yet but it is likely to be over £300 to balance so is there a cheaper or better way of doing this?

I am now looking at what is available to quantify vibration in an engine on a test bed. Recently Ford started to test the engines on the production line. Rather than them powering themselves they drove them through a shaft and sensed the vibration. They then added or removed mass from the flywheel to reduce vibration. Various hand held vibration sensors and many vibration transducers are available. Surely someone in China has come up with a set up like my wireless endoscope that cost £15.

Here is a view down the oil gallery of the block I am using which was from a half rebuilt engine that had supposedly been cleaned. Those black blobs are lumps of oil sludge and I did find something harder (probably solidified Hermetite) that would have blocked a small gallery such as to the rocker cover.
Amazing being able to see down the inside of an oil gallery. All galleries were pressure flushed with diesel, rodded and blown with compressed air backwards.

[Srenner]

Bolt the clutch cover to the flywheel, as that is part of the mass to be balanced. The friction disc is not as it obviously can rotate separately relative to the balanced mass. I typically pay to have the crank/pulley balanced as a unit, and the flywheel/clutch cover as a unit. I know this as "zero balancing" , so that if I need to swap cranks or flywheels, there is no need to rebalance the assembly.

My understanding is that bob weights are not used on 4 cylinder cranks as the journals are in-line. Bob weights are needed for dual plane v-8 cranks.

[Keith Clements]

Thanks Scott.
If I have the inclination I may try to work out if bob weights are needed on the Jowett engine. Whilst it is true that there is some amount of cancelling between 1 and 2 and between 3 and 4 and that reciprocating weight is counterbalanced by the webs of the crank there MAY be a dynamic moment that needs to be balanced out in the assembly.
I have already spun up the crank, flywheel and clutch assembly on the lathe and although I do not have a quantitative value of gm-cm or ounce-inches they do seem to be reasonably balanced.

One issue I will have on the next build is when I take a load off the flywheel to lighten it. As I intimated above, until I do the math, I will not know if this would upset dynamic balance of the assembly if the flywheel was balanced on its own after machining. My gut feel is it would not as the Jowett configuration is inherently balanced.

Last couple of days have been spent sorting out con rods. The 'set' in the engine being built up were found not to be a set. Started off weighing the piston + rod together and had a 54gm variance over the set! The reason being one rod was of a different variety so I went through all my spares and eventually found one with the same cast number.


I pressed the pins out and weighed all the bits separately. The pistons/rings/pins were all within .5gm. The now set of rods were now within 9 gm. After weighing the small and big ends using a WD40 can as a big end journal and an aluminium bar as a small end journal I weighed each end. There was 6 gm variance on big ends and 4 gms on small ends. So three stages of grinding brought the BE, SE and total all within a gm . Note some care was needed with the balance to obtain consistent readings. I found a layer of soft foam on the balance to allow the end to sit on helped. This is probably due to point loads upsetting the vectoring of the load from my makeshift pivot point. It helps if the pivot centre ( the aly bar) and the contact point on the balance are at the same level to create a vertical load without a sideways component.
I also looked at my other 4 sets of con rods from quite dead engines that I have acquired over the years which all exhibited mismatches of rods. It just shows how little care was taken previously by other owners or even perhaps the factory reconditioners.

I also checked the bore for size and ovality.


Oh and to answer a question asked of me by Alf Heseltine a couple of months ago as he was worried about using fan washers under the big end bolts. All the 4 set of rods taken out of these old engines had internal fan washers under them. I could not remember what I put under the Unibrako bolts in the engines I rebuilt. That engine is the next to be done. I use threadlock on the bolts, so perhaps a flat washer would be safer than the brittle fan washers.

[Srenner]

Take a look through your rods to see which ones have a small ridge that flows from the beam over the bump that houses the enclosed rod bolt and blends into the area above the serrations for the big end caps. These are a little stronger in the area where the rods break; where the beam flares into the rounded shape of the big end. I could dig one out and post a picture if I weren't so lazy.

Here's a link to a picture of the current approach to getting a lighter flywheel:
https://www.moss-europe.co.uk/lightened-flywheel.html.

The old way was to lighten was to hollow out the back. Easier to do, but need to be careful about getting close to the flywheel-to-crank mounting bolts. One of my Sprite flywheels looked like a boat propeller surrounded by a ring gear. Still seeing active duty after 20+ years. As the Jowett flywheel does not have counterbalances as part of the flywheel casting ( see a Ford 289 V8), all that needs to be done is to balance it after the machine work is done. With the clutch cover bolted on.

Nothing but fun!

[Keith Clements]

Just looked at the engine from the SA and the 3/8 BSF x 1+5/8 inch cap head socket screws were assembled without washers. Just ordered a stack more but they now seem only to be available in 2 inch length so will have to cut down. They are 12.9 tensile strength, probably double the original.

I have never experienced a broken con rod on a Jowett, although I did on my GTB Suby. The Suby explosion was caused by a worn BE bearing allowing the piston to hit the head. The car was being driven carefully on its way for a diagnosis of a ticking sound when it let go. The smoke screen temporarily blocked the motorway.

[PJGD]

Oval web cranks are heavy, so it makes sense to match it up with a lightened flywheel. Hollowing out the back is the right way to go while leaving as much mass as possible out near the rim so that you get a low total mass to inertia ratio. The Moss flywheel is quite radical!

Keith; I am interested in your low cost wireless endoscope - what is the make and what does it look like?

[Keith Clements]

http://powerspark.co.uk/tools-timing/borescopes/

[Keith Clements]

Also of consideration on the flywheel is the thermal capacity of the flywheel as it needs to absorb the heat of the clutch. Thus thermal stress needs to be considered as well as the mechanical integrity of something spinning at 8000 rpm and accelerating and decelerating to that speed in a few seconds. The use of spokes, cooling fins or a propeller shape to disperse hot air is worthy of consideration.

The flywheel I am putting in has already been lightened by simply taking 10mm off the front from the outside until well before it would meet the depression that secures the bolts. Taking anything off the back side (clutch face) from the inside out is governed by the width of the friction plate. I suppose one could construct a ventilated disc similar to a modern brake disc but that is too radical I suspect. One could also hollow out depressions on the front side whilst leaving a few spokes to support the outer mass and ring gear making the clutch plate fairly thin. This is also too radical. I guess the main question is how thin is it sensible to make the part in contact with the friction plate? One could then leave more mass around the ring gear.

I am still hoping to find a vibration meter that is priced within my budget. Enquiries to various piezo electric transducer makers have not yet yielded a result.

[BarryCambs]

Hi Keith

Just for interest, my Moto Guzzi V-Twin engine parts came back from the balancer a couple of weeks ago. They charged £200, including carriage, for the crank, flywheel, clutch cover, pistons and rods. Interestingly, our local and trusted engineering workshop wouldn't do it and it seems there are only a couple of places who will tackle V-Twins. Just need to find the time to rebuild it and see what it feels like.

Barry

[Srenner]

I have to disagree about leaving mass at the outer edge. The goal is throttle response. The whole idea of a lighter flywheel is to lessen inertia. Less mass means faster gain in rpm and better engine braking by not having mass maintain rpm.

For a long time, aluminum flywheels with a steel insert were considered the way to go. Turned out not to good for longevity as the bolt holes in the aluminum would distort. This brought about the radical billet steel flywheels as seen in a previous post.

Racing clutch packs moved to multi-discs in as small a diameter as feasible decades ago to lessen the effects of mass at a distance from the rotational center. Heat in the clutch pack should not be an issue unless the friction disc is slipping. In most of this type application, the clutches use metallic friction discs and this does eventually warp the steel inserts. Launching with this type clutch is like pushing a button: on...off. Very easy to stall.

I recall that a stock Jupiter flywheel weighs about 10 or 11 kg or about 24lbs. The lightest flywheel I have is 11 lbs or about 5 kg. No degradation in driving quality. Used it to drive to the Arctic Circle!