Thwaites 2 Ton Two Cylinder Dumper Running On One Cylinder

[awemawson]

Many years ago - (probably 9 or so) I bought a 2 ton Thwaites articulated swivel tip dumper powered by an electric start two cylinder Petter 'PH2'. I got it as we planned to move to a place where I needed to break up and shift over an acre of concrete 

Well that move fell through, however it came in very handy at my current place. One cylinder was always a bit reluctant to fire up until warm but I 'got by' for all those years.

Truth be told shortly after I got it, I exchanged injectors between the cylinders to no avail, and checked that both were squirting mist, but I was inhibited from further investigation as not only was time very limited, but so was access to the engine. The seat frame arrangement spans the top of the engine, and as well as the seat it also holds the air filter, the hand brake lever and the tipping hydraulic control. Just unbolting it wasn't an easy option, so it got ignored 

[John Rudd]

Is that you driving in the first picture?

[awemawson]

Now this machine has done some serious shifting for me, but it's been getting progressively worse over the years. I managed to convince myself that the front cylinder must be low compression. So much so that when I was at a 'close out' sale of a plant hire place and I saw that they had some PH2 barrels and pistons I moved them from their shelf to mine for future repairs. Well those barrels have sat on my shelf for five years 

As most of my heavy building work is now done and time a bit freer I thought I'd have a crack at the dumper. First issue was that as it's an electric start with no way to hand crank it, I couldn't easily prove the low compression theory. The main crank shaft is 1.5" with a 3/8" key way, and on the hand cranked versions there is a special handle that has a trip dog that engages with this keyway.

Trouble is that handle wont fit through the hole in the back plate of this digger , also it costs £44 

[awemawson]

Is that you driving in the first picture?

Oh yes - and that's my youngest daughter - whose now turned 30 and is married with two girls and another on the way 

[awemawson]

Now I don't need to crank start it, it has a starter motor, but I do need to turn it over in a controlled fashion to check the valves and compression, so I welded up a rough and ready handle.

I did start to broach a keyway in that collar with the Bridgeport slotting head, but my 3/8"slotting tool wasn't long enough, so I came up with a 'bodge' key comprising three 3/8" slugs of mild steel welded into suitably placed holes.

[awemawson]

So at least now I could manually turn the engine over, and darn me the compression 'seemed' good on both cylinders. Now I do have a petrol engine compression tester but it only goes up to 300 psi, so not a lot of good for a diesel engine.

Then I thought, if I open the exhaust valves on each cylinder in turn, crank the engine over on the starter with only one cylinder compressing, then surely measuring the current taken by the starter should give a rough comparison of the effort required and hence the compression.

The following readings are:

A/ Both cylinders decompressed (presumably least effort)

B/ Front cylinder compressing rear not

C/ Rear cylinder compressing, front not

[awemawson]

  Now how odd is that 

It is drawing MORE current when cranking WITHOUT compressing than when the valves are closed - I can't explain that - it may be that the engine turns over faster, but it didn't seem to.

However the comparison between cylinders showed that they weren't miles off each other, but I do find those readings EXTREMELY odd 

[awemawson]

So I decided that probably we have an injection timing or volume issue. The pumped volume to the 'faulty' cylinder seemed less than the other, but it's a rather subjective judgement.

All the timing and volume settings are on the right hand side of the engine at a low level, with access totally blocked by the diesel tank. But before that the seat base had to come off. The air filter and hydraulic control weren't too bad, but the hand brake lever and cable were silly! The adjustment screws into a plate that was welded on under the seat mount, and without stripping the cable out all the way from the front axle (via the articulation pivot) it wouldn't rotate to unscrew. At this point I decided that it was a stupid design, and cut through the welds with an angle grinder and will re-engineer it when the time comes to put it back together.

The tank is used as a structural element supporting various horizontal bits of chequer plate so the left hand 'mudguard' (6 mm plate) and a cill plate (more 6 mm plate) had to come off before the tank fixings were accessible.

(Pictures are slightly in the wrong order)

[awemawson]

So now at last at least I can see where all the twiddly bits are without standing on my head and using a mirror 

[awemawson]

There is an injector pump for each cylinder, and their 'volume racks' are linked to move in synchronism to give equal volume to each. There is a governor in the bowels of the engine that moves the racks against a spring which at rest presses the rack end against an 'overload stop'.

The setting up requires that the racks are first synchronised so that a calibration mark on each is 13/16" from the spot face on the pumps - this is set by adjusting the length of the link rod that joins them. However this setting needs to be 'at rest' when the rack is limited by the 'overload stop'

Problem is the overload stop is hideously worn and flopping all over the place

[awemawson]

It is basically a little hinge, one part fixed to the end of one rack, and the other when folded down carries an adjustable bolt that presses against the 'overload bracket'. The hinge is a clevis and tenon, both of which were very worn

(This 'hinge' is folded up when cold starting)

[awemawson]

Now the clevis part is available for £27 but not the other bit, but why buy when you can make?

I decided that there was enough 'meat' left in the old clevis to open the slot up to 6 mm, re-bore it's pivot hole (also now 6 mm), but to make an entirely new 'tenon' to suit the new size of the slot.

All then I needed to make was an over size pivot pin (6 mm x 18 mm) - I also reproduced the 'wiring hole' that lets the adjusting bolt be wired to the flappy bit of the hinge to stop it vibrating loose. I just LOVE drilling 1.5 mm holes 1/2" deep - not !!!!

[awemawson]

So when the rain stops I need to delve further into the engine - there are two covers, one below each injector pump, that (should !!) reveal the adjustment that sets the timing of the injection for each pump.

Essentially there is an adjuster that sets how high the pump plunger sits on it's cam follower. By removing the output valve on each pump and setting the fly wheel to the correct point for injection, the adjuster can be tweaked to 'just' stop fuel flowing (gravity from tank) - apparently this is called 'spill timing' and is common on these sorts of engines.

But first I've got to master and sort out all the levers and fixings that are mounted on the rear cover

[appletree]

Until my dad died 22 years ago we had a plant hire family firm which we started in 1972, we had lots of Dumpers, Pumps, concrete mixers with Petter singles and twins, Lister singles twins triples and 4s. I was going to tell you about the setup on PH series engines. As I thought i’ll tell him XYZ I further read the post and what I was going to say was in front of me. Invariably there is a strong cylinder regarding starting, seems a bit odd as nothing obvious is standing out yet.
 

[DavidA]

You could make up a simplified version of the differential compression tester.

Remove the injectors.  Make an adapter that will allow you to connect an air line to the cylinders (one at a time) then, with the piston held at top dead centre, pressurise the cylinder.

You can then listen for any air leaks from the inlet and exhaust ports. Also listen for the sound of air getting into the crankcase past the pistons.

Just a thought.

Dave.

[appletree]

There is an injector pump for each cylinder, and their 'volume racks' are linked to move in synchronism to give equal volume to each. There is a governor in the bowls of the engine that moves the racks against a spring which at rest presses the rack end against an 'overload stop'.

The setting up requires that the racks are first synchronised so that a calibration mark on each is 13/16" from the spot face on the pumps - this is set by adjusting the length of the link rod that joins them. However this setting needs to be 'at rest' when the rack is limited by the 'overload stop'

Problem is the overload stop is hideously worn and flopping all over the place

Perhaps not immediately obvious to our readers the overload stop is disengaged for starting, it then drops back into place once the engine is running and the governor takes over.
The purpose of the device is to prevent over injection when the engine is under large/excessive load thus preventing over fuelling .

[Manxmodder]

  Now how odd is that 

It is drawing MORE current when cranking WITHOUT compressing than when the valves are closed - I can't explain that - it may be that the engine turns over faster, but it didn't seem to.

However the comparison between cylinders showed that they weren't miles off each other, but I do find those readings EXTREMELY odd 

Andrew, what I believe you may be witnessing here is an atmosperic engine braking type of thing.

Decompressor engine brakes,usually only seen on 2 strokes, work on the principle that the piston has to work hard to draw air in through a small orrifice under only atmospheric pressure hence creating a partial vacuum in the cylinder.

On the subject of testing the cylinders for compression,why not make up a simple compression tester that replaces the injector with a suitable pressure range gauge fitted.
......OZ.

[RussellT]

If you want a comparative reading rather than an absolute figure you could use your petrol engine compression tester.  If you used an extension pipe between the tester and the cylinder you would increase the unswept volume and decrease the pressure reading to a suitable amount.

I'm following with interest.

Russell

[awemawson]

Thanks chaps but at this stage I think that the compression issue is a red herring. Given fine weather tomorrow I hope to go through the setting up procedure for the injection timing and metering.

Visitors today, but I did manage to make up a 'swan neck', that screws onto the output of the pump being set up, and acts as a 'stand pipe' to let you observe the 'spill cut off point' more accurately as you tweak the cam follower tappet adjuster.

The piston of the injector pump rises, and at one point covers the 'spill port' at which time the pressure starts to build and fuel is injected. The piston has an angled cut in it that, as it is rotated by the rack alters the effective point at which cut off occurs and hence changes the volume squashed though the injector. So it is important to set the rest point of the rack first, and that depends on that little brass 'overload stop' I re-made the other day.

[awemawson]

Well not the best day in my life 

It all started quite nicely. I went through the procedure carefully to set the rack synchronisation and then went on to the  injection timing using the 'spill' method. I started by cleaning up the marks on the fly wheel - it carries TDC marks for both cylinders (as they are 180 degrees apart) and also for 24 degrees and 28 degrees before TDC as it seems 24 is for engines running up to 1650 rpm and 28 degrees for engines running up to 2200 rpm. No plate on mine so I decided to use the 24 degree setting.

Starting with #1 cylinder I removed the injection pumps one way valve as instructed, and installed the 'Swan Neck' slackened off the adjuster as per the book, set the timing mark bang on and then raised the tappet screw until the flow stopped - all went apparently as it should. I took the precaution of taking a picture before and after to be able to compare.

[awemawson]

As you can see it took about two turns to set it at the spill point, which was a very definite setting - I went past and approached it again several times to get it 'spot on' and I can see why they suggest the use of the swan neck rather than just letting it spill over the top of the fitting. It slows to a series of drips which then can be tweaked to no drip.

I then put the one way valve back in the pump, reconnected the injector pipe, and thought it wise at this point to prime fuel though. Stupidly I just flicked the starter motor to spin the engine and there was an almighty BANG and bits flew in all directions   

Looking at the poor old pump it had disintegrated - blowing itself off it's mounting flange. At this stage I don't know if perhaps the one way valve mis-seated and it hydraulically destroyed itself. Or if perhaps having raised the tappet two turns the pump piston contacted the top of it's cylinder and mechanically broke it. I'll certainly need to find out before another pump goes on

I have to say that as I was doing the 'rack synchronisation' settings earlier I'd thought that the whole pump / rack set up was rather worn and sloppy, so perhaps this is a blessing in disguise.

Anyway I've dug deep into my pockets and ordered up a pair of injector pumps, so play is suspended until they arrive (so are most other activities like food and drink as they were not cheap  )

[RotarySMP]

Well illustrated report thanks Andrew. Shame that the injector pump has left us. Maybe it is the sort of building site machinery which really prefers never to get maintenance

Mark

[Manxmodder]

I guess that if it's mechanical destruction and the plunger piston has made contact with something then there will be some marks internally to evidence that.

Does your flywheel bear two sets of timing marks? On some of the Petters I,ve seen in the past there is a  second set of marks duplicated on the other side of TDC.

These other timing marks are for a counter rotating engine build spec......OZ.

[awemawson]

I think most Petter flywheel have the marks on each side of TDC as depending how they are assembled they can run clockwise or anticlockwise (with the right internals) No real witness marks evident showing mechanical contact. The body of the pumps are cast aluminium so they've been replaced before probably. I think Petter had later bits made in India.

Spent the afternoon drawing up gaskets in Autocad for the side covers. Could have tapped them out but access isn't ideal and the Laser Cutter makes such a neat job of things - far better than scissors or knife 

(Did I say  -  I LOVE my laser cutter  )

[petertheterrible]

Excelent post.  Only things I would like to add, watch out for 'over spilling'.

Only thing the PH manual does not state is, in setting spill timing before tightening the follower setting screw is to turn engine by hand, preferably not with the crank but just the flywheel.  (Stated in manuals like that of the SR strangely) 

Each pump must be checked for 'overspill', if not you will destroy the pump on the PH. (Experience gained the costly way)  (On some other models bending the camshaft).

Slight 'overspill' will cause the pump head to rise (inside pump self) causing diesel in the crankcase.

The most likely thing when these twins and threes aren't performing equal on all cylinders is dirt in the pump and injector.
Dirt causes the pump to spill differently as the piston is obstructed and when the injector is also dirty, incorrect injection isn't detected by eye.

Personally I just pull the injectors an test them.  A lot can be determined by just testing injectors which are usually easy to locate and remove.  These engines perform nearly with no compression and always amazes me how easy they start compared to something like a Wolseley.