ABV intake dyno evaluation. Farewell. Toyotec's Golf 3 200ps VR6 'Der Ankor'.

C'mon Ed... updates on de Anchor!

 

Sold!

 

Noooooooooooooooooooooo!!!

You scared me in that beast once upon a time

 

Part 2. Shed clean out and head prep for porting.

As indicated above, the car my family called "the soldier" or "Der Ankor" was sold last year. I do miss the holla of the slicky VR6 motor. However before it was sold, modifications were done to the original engine that coaxed power to 200bhp and 212lbft of torque, with the snappy response as standard from an engine map and a lightened flywheel.

Before taking up the voluntary roles here at Club GTI, I had planned to update this thread and so I will attempt to bring you all up to speed now.

Following the stripdown of the spare 12v cylinder head, we continue with a shed clean out.

Dec 31st 2012
We continue. I needed my shed back so decided to have a little clean out.



...well actually a very big one that involved the whole back yard!



Once the scrap, like lesser 8v valve heads was thrown away, these remained worthy of keeping



Now the head is at the table ready for dissection!



During the clean up, I even found some useful bits for my 16v project as well as potential 2.9 rebuild work.



Tomorrow is New Years day and work will proceed.

January 1st 2013 - A new year with old challenges.

2013 started with a bit of a bang!



With the celebrations out of the way, I got up this morning returned to where we left off on Monday i.e. stripping the ABV cylinder head.

As this head does not have valves shrouded by a combustion chamber, being diesel like in configuration, it could be placed flat on the table and the valve collets could be bounced out with a hammer. If you are not experienced with this sort of job, to prevent potential damage to the top of the valve stem, in hindsight the preferred method is to use the valve compressor

First the plugs were removed.





Then the hydraulic lifters were removed



With the head flat on the table, an old spark plug socket, along with an old 3/8 drive extension, was placed concentric on the valve spring and the 'bumped' until the collects were freed. The socket acted as a catch for the collets that would inevitably make they bid for freedom somewhere into the abyss that is my shed!



Here are the collets now freed from the valve stem.



And now removed with the aid of a magnet.



All the lifters were placed in container as they appear in the cylinder head.



The valves were placed in a similar order in a makeshift holder, made from a cardboard box.



At least 33 pounds saved in "cylinder head organizer" fees.

With the head free of valve gear, it was time to start cleaning the carbon off.









Once the majority of crud was off, it was taken outside and given a once over with the pressure washer.



For ports, excess carbon build up was removed up with a wire wheel.

N.B. the valve spring retainers were not removed as were the old valve stem seals. To be done on re-installation of valves,

We shall continue this write up with next stage - Porting



Stay tuned, Stay optimised.

 

Last update photos aren't showing.

 

Now fixed all. Needed to upload direct from PB.

 

I had the same problem when I copied the pictures from Facebook to a thread, GVK said he could not see it. So in the future will just copy from photobucket.

 

Part 2 continued 12v VR6 head porting.

VR6 head porting.

Cylinder ports are developed and designed, to flow a quantity of air to support swirl or tumble towards homogeneous combustion. In the ideal world of CAE modeling, ports shapes can be simulated until the design can demonstrate near to complete combustion for OEM targets. Once these targets are met the design is prototyped, tested, validated and eventually the final sign off is then put into production.
Previous work from my aftermarket dyno testing at Garage Streamline, we know factory cylinder heads, with their many inconsistencies, can meet OEM power and torque targets. With a bit of targeted ECU tuning even exceed these conservative measurements.

In the 12v VR6 engine, ports are offset to the side of the cylinders to prompt a swirling motion in the cylinder chamber. Ports also have different shapes depending on if the runner is long or if short.
The 12v VR6 valve seat diameters, on both inlet and exhaust, are quite optimized compared to EA827 16v and 8v variants.

Looking into port throat after a wire wheel was used to take out any carbon build up.



All ports did have mismatching. The areas of optimization on the ports were:

Long radii port to seat mismatch (all ports)



Short radii port edge (all ports)



Exhaust port - step on long radii from mass machining.



So the porting gear came out and work progressed with the carbide cutter first.



Before,





After,





Care had to be taken when blending the lower part of the seat to the port. As the die grinder was variable, the cutter speed was set to 7krpm and lots of WD40 was used.
Here the seat is being cut down to reduce the step and be flush port surface as illustrated by the freshly cut alloy.



This area requires further seat grinding to blend flush with the port.



With all seats and throat work concluded roughly, the chaff was washed off.



We continue with intake manifold to head port matching then final flapping.

Stay tuned stay optimised

 

are the valves in this head standard size? and same as 2.8 head?
i only ask as there doesn't seem much space to go bigger, yet you can buy +3mm intake and +2mm exhaust valves for the 12v VR6 - and i can't see how they'd fit!

 

do you know if the increase intake runner length on VR's mean one bank of cylinders has more / less power? or is it compensated for by the intake manifold?

Like the porting - that is effectively what I have done on my 16v head. removing the steps from the seat into the throat.

 

Yes the ABV head here, has the same valves as the AAA engine. The LCA on the cams is different.

 

We will get to that later. More an increase in plenum volume

 

Part 3 Port matching and overheating investigation

Manifold match porting begins.

On Thursday, Jan 1 2013, means from 07:45 I start my first working day of the new year.
To prepare for my impending lack of time, On the New Year's eve, I spent some time searching for engineer's blue to assist in marking out the areas in the ports and inlet manifold to be matched up. The exhaust side was left alone as the mismatch, where the exhaust tube is a larger dimension to the exhaust port, is great for anti reverse flow, when pulse tuning. As I could not find anywhere that sold engineer's blue over the shelf, I took the old inlet gasket, attached some screws to secure the gasket to face of the head or lower intake and began porting to match the port shape using the gasket holes as the limit. A better job could have been done with engineer's blue and a scribe.
The roof of the port was where the most time was spent. There were some steps on the floor of the port that were also blended in. My camera went down while this operation was carried out, so only have a few pictures for before and after with a first cut.

Standard ports




Opened up ports




Intake lower manifold standard




Matching inlet port.



Std head port vs modified first cut.




Same for the lower manifold.



Further head porting to continue in the next update.

Overheating concerns.

It was observed over long soaks or periods where the car would not be used, on a cold start, the engine would misfire on one of two cylinders for a few seconds and clear into its usual hum. As spark plugs were changed for new, there was no problem with the plug leas, the coil pack was fresh and there were no fault codes, I was of the belief that the cylinder head gasket had ruptured by a water passage and this path was letting water into a cylinder or 2 when the engine was off over a soak. On my journey today, Matt of Kilburn Garage, lent me a coolant testing kit.



My plan is to check the Ankor's hardware to ascertain were the water is going, incase it is not limited to the worn engine.

With the car cold and soaked over night, I proceeded to set up the test by connecting to the adaptor to the radiator expansion tank and pressurizing it normal operating pressure of 15 psi. I would expect over 10 mins to loose no more than 1-2psi attributed to the tool.



The system was pumped up and timed.



After 10 minutes had elapsed, the remaining pressure was checked.



That does not look good 5 psi loss in 10mins. There was no visible signs of water loss on any of the cooling system. This pointed to a damaged head gasket.

So check the tool was working properly and was not introducing an error, the same test was carried out on a new 2012 Focus.



After 10 mins only a 2 psi loss was recorded.

As the cooling system was being pressurised with out starting the engine, when there was an attempt to start the engine, it was very hard to start and misfired longer than a few secs before clearing up. I was convinced the head had to come off which meant finishing the modified head.

We will continue with flapping up the head ports in the next update.

Stay tuned...Stay optimised.

 

Part 4 Completion of head porting and manifold matching 06-01-13.

Just to recap, the cylinder head work that had taken place, focused on port matching and blending the throat into the seat area. It is strongly believe these areas carried a penalty to head flow for this type of cylinder head.
As this head has ports offset to create swirl, no modifications would be done to the port contours generally. In fact in long ports to cylinders i.e the intake 1,3 & 5 and exhaust 2,4 & 6, were not tweaked just after port entry or before throats.
In addition further work could have been done the reduce the seat thickness, including a 3 angle valve job.
As the head gasket on the vehicle was determined to be blown, there was pressure to get this modified head back on the vehicle. With the valve guides to be in serviceable condition, professional machine work to gain more flow around the valves was not an option.

With 3 ports already prepped with the carbide cutter, the remaining ports received the same treatment.

Standard



Shaped with the carbide cutter.



First cut blending done to the throat.



Port and throat flap wheel finishing.











Head porting complete.



Attention then turned to the intake manifold and this was completed to have the same finish.

It was observed there was a step on runner holes that married up to the upper manifold.


This was first prepped with the carbide cutter and then flapped off.








Next, the valves were decoked.



A drill was used to hold each valve and the carbon was removed with the flapper cloth.



Much better!



All the valves cleaned up and ready for lapping.



Parts ready for assembly.



Next Steps:

• Choosing the cams. Started 30-12-12

• Purchase Pastigage. Done/Found
• Inspecting ABV engine bearings and pistons. Due to commence
• Honing the ABV block. TBA
• Purchasing the rings. TBA
• Test Der Ankor for coolant leaks 2. Due by 09-01-13
• Head prep and matching porting. Done 06-01-13
• Valve measurement and prep for relapping. Due by 09-01-13
• Head assmbley. Due 09-01-13
• Match M3.8.1 J338 thorttle module to 2.9 inlet manifold. TBA
• Engine reassembly. TBA
• Purchase sundries. In progress

We will continue

 

Part 5 Coolant leak down 2, Valve seat lapping and leak check, complete head assembly

We continue the works on my Golf MK3 VR6.

Recapping the original intent and future work

Just to recap, I had hoped to build a 12 valve engine to supersede the performance of a stock 2.8 24v engine, that at the time was costly and I did not have the gear to remap the engine.
The plan was to use and refresh a 2.9 ABV engine, picked up due to a deal. Of course this would be done while the current car was working normally, which now seemed to develop a problem. A trackday was also booked and the car was used as a daily commuter, putting pressure on the exercise. Therefore there maybe changes to the plan as the story develops.
We carry on.

Engine coolant system check 2

Following the coolant leak test, I was convinced the pressure loss seen in 10mins on the Ankor, without leaks, pointed to a blown head gasket.
To confirm the was definitely the case, the following night the system was pressurized to 16psi and left overnight and the level of the expansion tank noted. On returning to the car the following morning, the pressure was zero and the water level had dropped slightly. There were no external leaks and that suggested that the water was in the cylinders.
So the engine was started and sure enough it misfired on 1-2 cylinders and then cleared up after 40secs of elevated engine speed at 2000 rpm. This confirmed a blown gasket was causing the water loss. The water was believed to be leaking into the cylinder, when the coolant pressure was higher than the combustion chamber.
This put pressure on the project to finish. Would the ABV engine turn out to have minimal wear, to allow the entire assembly of modified head and refreshed short engine to be fitted to the Ankor in time for the trackday? We will see...

Lapping in the valves and testing for leaks.

Following our last segment, it was night and I really needed to push on with this project.
So a makeshift stool was taken from outside and put into the house to allow me to take on the boring task of lapping valves while watching TV.



And no the wife was not impressed lol



All valves were ground with first the coarse paste to quickly break off the carbon and then the fine for a polish.
Shown here after using the coarse paste.



Eventually all 12 were done and the head was ready for assembly. That was it for that night.



The next morning, before refitting the valve springs, a check was done to see if the lapped valves were seating.
This involved ensuring both valve and valve seat were cleaned of any grinding paste, shut against the seat and pouring water down the port to see if any would escape at the valve.

NB Following a recent offline discussion, it was advised to back cut the valves as well as carry out some work to the seats. Admittedly in 2014, as there was no ready means to work the back of the valves and machine work would have added cost to this exercise, it was decided the limit modifications to the seat and port mismatching. The dyno WOT delta of this effort will be shown nearing the end of this thread.



This picture shows the test being done to an exhaust valve which checked out all good.



With all 12 valves passing this test and the lot sprayed up with WD40, it was time to assemble the cylinder head.

Complete cylinder head assembly

When the head was stripped, I deliberately left the old stem seals as they kept the lower spring platforms in place. Time to bring out my little friend



Within minutes, 12 old hardened seals were out!







When the new seal set was opened and there was a problem. Spot the problem?



It would seem the Wemebley branch of ECP managed to open the bag and replace one of the 12 stem seals with a seal from another type of engine. It later turned out I have an exact set of 20 of these green stem seals, meaning the green seal seen in the pic is from a 20v set. This halted the assembly process till the next day . Not impressed

Later on the next day (after 5pm Sunday) another set of seals were purchased from a difference supplier ( Lloyds aka Midnight Motors open till 10pm). Once the new set were in hand, the stems seals were fitted using the tool as shown in the pictures.








With the new stem seals firmly in place, the head was now ready to have the springs and valve collets fitted. So out came the valve spring compressor, magnet, small flat blade screwdriver and hair grease!



Spring was compressed so the grooves could be exposed.



Then the magnet was used to place the collet into position. I later revised to the screwdriver and grease as this proved to be more maneuverable.





Valve springs all fitted, the tappets were retrofitted.



The head was not fully built and ready to be fitted to the short engine.



That is it for now.

The next steps for this exercise.

• ABV engine stripdown and clearance measurement.
• AAA/ABV camshaft types, LCA angles, aftermarket suppliers and short listing
• Completion of collecting and purchasing sundries.
• Based on finding on ABV base engine, switch the plan B where the original AAA engine in the car, will be used with modified head.
• ABV intake manifold modification to work with later adaptive throttle body footprint.

We will continue in the next segment.

Stay tuned stay optimised

 

ABV 2.9 12v short engine teardown and post-mortem

It has been a while since I last updated the works done to my past Dragon Green MK3 VR6 aka der Ankor. I wish to summarise this chapter as times have moved on since I owned that car.

Recap from post 137.

As a reminder, these were the tasks and status remaining somewhat captured in the last post.

Choosing the cams. Started 30-12-12 now postponed as DRC link gone cold.
Purchase Plastigage. Done/Found and used in 2.9 bearing inspection below.
Inspecting engine bearings and pistons. Commenced 15-02-13 found concerns.
Honing the block. Found concerns mid bore on some cylinders 20-02-13.
Purchasing the rings. Delayed until further notice.
Match M3.8.1 J338 throttle module to 2.9 inlet manifold. Work commenced 09-03-13
Engine reassembly. Delayed until further notice.

ABV 2.9 short engine teardown and post mortem

In mid February that year, strip down inspection began on the 2.9 ABV short engine.

Removal of the OE 5.5kg flywheel



Lower timing chain with cover removed.


Oil pump and sump


Engine internals had a golden brown appearance, suggesting regular oil changes over the 187K miles. lower timing chain and guides/tensioner in excellent condition.



As the bottom end bearings were now fully accessible, the main bearing caps were removed and plastigage was used to measure the crush. This crush/clearance was referenced to vw workshop manual specifications for the VR6 AAA/ABV engine. The same procedure was carried out for the connecting rod big ends before removing the pistons.

All main bearing shells were of a similar appearance as seen in the picture below.



The crankshaft bearing surface looked in good condition as well.



Time to gather clearance data using Plastigage.



The bearing caps, complete with shells, were retorqued with a small strip of Plastigage and removed to measure the crush. This level of crush/clearance compared with the markings on the Plastigage wrapper.





These measurements were then recorded.


Conrods shells measurements being taken.


Unfortunately conrod shell 4 had some damage



As did the shells for conrod no. 5.



The crankshaft surface was not damaged but new conrod shells would be required for a rebuild.

A table of the final clearance results and limits was complied and can be viewed below.

	Conrod Clearance		Main bearing Clearance
	New 0007"-002" - Limit 004"		New 0007"-002" - Limit 004"
1	0.0022				1	0.0025
2	0.0023				2	0.0020
3	0.0025				3	0.0022
4	0.0035				4	0.0023
5	0.0032				5	0.0020
6	0.0020				6	0.0025
					7	0.0020

Next steps, piston and bore inspection.

The most of the bores were in good condition, with the crosshatch still visible from TDC to BDC of the bore.


To ensure all 6 bores were wearing uniformly, a bore gauge was used to measure the bore in line with the piston pin and perpendicular at positions from 11 mm from TDC (A), middle bore (B) and 11mm BDC (C). The results were then tabulated.





During this procedure, bores in 2 cylinders were found to be worn close to the limit (bore measurements were not complete when this was discovered . Rather than rebore to an oversize diameter >82mm and as the conrod bearings 4 and 5 were scored, decision was made to obtain a std 2.8 block and rebore from ~ 81mm to the 82mm required for a 2.9 engine. This action would keep costs to a minimum by allowing reuse of the 82mm pistons. It would also mean a time delay in getting the 2.9 engine ready to fit to the vehicle in time for a track day at the end of March. As a result emphasis shifted to rectifying the vehicle with the existing 2.8 engine in preparation for a track day.

The original plan was amended as follows:

• Replace head gasket on AAA 2.8 engine with a MLS component.
• Fit ported 2.9 head and reuse 2.8 std cams.
• Modify and fit 2.9 inlet plenum.
• Fit new FCP590R pads (DS3000)
• Fit Federal tyres on painted BBS Solitude wheels.
• Double check suspension settings

In the next segment, I drive the Ankor, after being parked up since Christmas, to Garage Streamline to replace the old fibre gasket and fit the modified head.

Stay tuned...Stay optimized.

 

Part 6 - Rectifying the original 2.8 12v motor.

Rectifying the original nearly 180K 2.8 12v engine

With the decision to delay the 2.9 engine build confirmed and Der Ankor required to take part in a track day, plans were made to take the vehicle to Garage Streamline and carry out rectification of the what was expected to be a damaged headgasket. Before setting off, the water level had not dropped since pressure testing the cooling system so the car was cold started. As expected, based on the results of the leakdown test, there was heavy misfire that would clear after 30 secs or so. With the roads clear the journey to the garage was without incident and removal of the cylinder soon commenced.

Original head being removed




Removal of the timing chain cover revealed this!



The long chain guide (part # 021109513) was broken after many miles of daily use, long hauls to Europe and trackday punishment, yet the car still ran as normal.

With the head removed, root causes for water loss/cold misfire could be investigated. The engine was originally fitted with a Victor Reinz fibre headgasket and first glances suggested everything was as any other high mileage engine.



However closer inspection near cylinder 1 revealed this.



It was quite evident water was getting into the combustion chamber as the coking from head on cylinder 1 was part cleaned off in the region of the head gasket damage.



With the root cause on concern identified it was time to prep for reassembly.
It was also decided to run the modified head that was prepared for the 2.9 engine and later on a modified 2.9 plenum which was larger (more on the plenum later).



The expectation was, the use of the modified head, larger plenum and track day tyres should add up to enough grunt on track as just adding a performance cam alone.

An Elring MLS gasket (401.820 ) was purchased on ebay for a grand bargain of 25 quid. It is believed by the interent these MLS gaskets increase compression by .5.



From previous measurements, it was observed the thickness of the the OE 021103383N MLS is 1.65mm. There are plans to compare this thickness with the fibre gasket that was removed.

The deck of the block was cleaned up and the MLS gasket put in place.





In part 1, picture was displayed with showed a stripped exhaust stud. Theo of GS used an extraction tool and removed.





The head was refitted, with timing and upper guides all phased with the crank correctly and torqued down.



Water topped up and the engine fired up into a smooth running VR6!



Next Steps:

• Replace head gasket on AAA 2.8 engine with a MLS component. Done 25-02-13
• Fit ported 2.9 head and reuse 2.8 std cams. Done 25-02-13
• Modify and fit 2.9 inlet plenum. At metal fabricators 11-03-13 complete by 15-03-13
• Fit new FCP590R pads (DS3000). Ordered from Badger 5 10-03-13 due in today 12th March
• Fit Federal tyres on painted BBS Solidtude wheels. Theo @ GS TBA
• Double check suspension settings - Gurds driven 10-03-13, Yaash Motors once over TBA
• Compare fibre gasket thickness to confirm actual CR. Done in this thread

In the next segment, the engine springs a leak, I measure throttles size differences between Bosch M2.9.1 and M3.8.1 and the intake manifolds/plenums are taken to Basildon for some reworking.


Stay tuned...stay optimized

 

Swapping a post 1995 ABV inlet to a 97 AAA engine. VR6 OBD explained.

2.9 and 2.8 intake system investigation and rework

The intake manifold on the 2.9 litre ABV engine has a larger plenum compared to the system fitted to 2.8 engines. In the case of the ABV that was bought, it was run on a Bosch M2.9.1 engine control system. This meant unlike the later control system fitted to der Ankor, it had a Pierburg throttle body and separate provisions for vane type idle speed control valve. Pictures show the ABV inlet system has more idle support components to work with the older M2.9.1 control system. This system does not have adaptive idle control as seen in the later post 95 system with the M3.8.1 controls.

Older ABV inlet system for M2.9.1 controls vs. later AAA inlet system controlled by Bosch M3.8.1


The intake manifolds from both engine variants look pretty similar from the when viewed as in the picture.

ABV M2.9.1 LH and AAA M3.8.1 RH. a


When the manifolds are flipped over the difference in plenum size is clear.

ABV M2.9.1 LH and AAA M3.8.1 RH. b


The throttles were also inspected and differences were found in the parts in hand.
The Pierburg throttle plate from the 2.9 ABV was measured at 64mm (Similar size and TPS to an ABF throttle body).









The VDO J338 throttle module had a throttle plate of 70mm, so in fact bigger than the Pierburg TB ( These can be great donors for 1.8T systems that use 54mm throttles and M3.8.3/(5.9.2 <- US OBD2) controllers.







This busts a myth that "OBD1" throttles (Pierburg) are bigger than "OBD2" (VDO) in the same way OBD2 controllers i.e. the M5.9.1 VR6 ECU was never applicable to Europe!

The Ankor uses a 'Euro' M3.8.1 control system to control its AAA engine. This Bosch controller is similar to ABV engines found in the Golf 3/Passat VR6 syncros from Aug 95 onwards. If a later post 95 donor 2.9 vehicles were available here in the UK, the intake system would have been a straight swap. Unfortunately the ABV manifold that was in hand was removed from a Corrodo Storm that was controlled by an older M2.9.1 system and the throttle foot prints were different between the Pierburg and VDO units as shown by the gaskets and throttle flanges.

ABV M2.9.1 LH and AAA M3.8.1 RH. c


ABV LH and AAA RH d


Also ABV Pieburg throttles 0211333061J use M8x25 Allen head bolts ad AAA VDO throttle module 021133064A use M6x45 Allen bolts

So the manifolds were referenced at a key point.

AAA


ABV


With the manifolds measured, they were taken to a local engineering center near my work to be integrated.
I will follow up on this in a later post.