View Single Post
Old 05-24-2016, 05:18 PM   #41
aristg
Board Member
 
aristg's Avatar
 
Join Date: Sep 2013
Location: Greece
Post

The story continues from post #2

UPDATE 24/5/2016
  • The new exhaust manifold did the trick ! Max temps seen are about 950*C on the turbine matching the reading on the first cylinder (close to the exhaust port). The EGT temps are now responding fully in alterations of fuel timing and the car feels that produces much "easier" more power ! Got it up to 2.0 bar of turbo pressure with no knocking or whatever. Seems that it can do more, but this will be discovered at the dyno session. End of EGT drama !!

  • Fuel timings at 2.0 bar are a bit abnormal: they don't seem to be in a logical linear way, but much more than expected. This meant that the stock fuel line + stock k-jet pump (did not have the 044) + stock fuel rail are not enough to feed high power.

  • The parts for upgrading the fuel delivery system were: NUKE fuel rail, Aeromotive 10 micron -10AN High Flow filter, Walbro 455lph E85 in-tank fuel pump, -8AN feed line, -6AN return line, several AN fittings and SS lines to weld. Both of the "old" pumps were deleted from the car since the Walbro outflows most of the pumps on the market !



  • Also upgraded the fuel hanger for the tank. The stock metal feed line (6mm ID/8mm OD) was replaced with an 8mm ID/10 mm OD SS line with a -8 AN fitting welded on its end. The return line from the flange up was also upgraded to 8mm ID/10 mm OD SS line where from this part i bypassed the smaller 4mm ID useless tube on the hanger and guided it to the bottom of the tank. The metal fuel line that holds the fuel level sender was not at all harmed, this is why i tried to increase the return flow through the bypass line. On the return line a -6 AN fitting was welded. I also tried to do all the electrical connections outside of the tank, so i guided the two FP cables through the original connectors which i drilled wider, as well as the fuel sender's wire, which also guided through the original connectors. Then all these were extra sealed with a gasoline resistant 2 component glue.

  • I wanted to have some fuel pressure readings in cabin, but i did not want to move the mechanical gauge, since i should have a fuel line also in cabin. So i bolted a second oil pressure sender on my Aeromotive FPR's 1/8 NPT port, and through an I-0-II switch connected it to the oil pressure gauge. So now i have a multi-gauge actually, that changes its readings flipping a switch ! A very easy 1-wire from each sensor to the switch and one from the switch to the gauge mod !! Just connect the G port of each sender to the switch .. and the switch to the gauge ..



The in-tank assembly before modifications



Original cable routing / sealing



Rusted Hanger Top



Got Better after 2 days in vinegar !!





Hanger mods



-8AN Feed / -6AN Return





The NEW final result


... go go ... Dyno Session ... !!!



UPDATE 5/7/2016

The 1st long dyno session is over which revealed many problems of my setup. Here is (not the best) pull. Note that these are numbers at the wheels.




The best pull got 373 whp by stretching the timing at about 27* (!!!) and hearing some knocking (actually 2 seperate knocks)

The problems:
  • The KJet filter box proved to be VERY restrictive, despite having a K&N filter in it. It could not provide enough incoming air to keep the boost constantly over 1.6 bar. The difference was amazing when it was removed and the turbo worked without a filter..

  • The intercooler proved to be insufficient and restrictive for these numbers. I need to get a bar and plate one with slightly bigger core. I think a 6.5x30x60 (cm) or a bit bigger can fit and provide some more cooling ability and better flow.

  • The turbo proved to be small for this whp number. Hence the ripply power graph, and the loss of power after it reached its top speed (the turbo). Since it was overspeeding the efficiency dropped below 55% and it started "boiling" the compressed air over 60*C. Also the top hp number is at 4500-5000 rpm and then it drops, which is not logical, showing that this setup has reached its limits. I will have to upgrade to GTX3071R or GTX3076R. Also i need to upgrade to a TS hot side, since i have this kind of manifold, with bigger A/R than .63 taking advantage of the TS properties. I believe i can get another 100 at least whp from this setup! Will find out soon

  • I need to play with the cam timings a bit to see if the power-torque band moves up or down to help having a DD car.

  • Next dyno session has to be in winter. About 33*C is not good at all for this kind of tests !!



UPDATE 15/10/2016

After having solid data from the dyno session, i decided to change a few (a lot as it proved) things to correct the max power performance of the car. At this stage improving stuff is actually much more expensive than before and require much more detailed research in things i have never thought of, but what the heck, i just can not stop there !! The new and final target is 500 whp so i just need to gain about 130 more whp. Hopefully these mods will take me there..
  • Changed over to a new hand made front mount intercooler, as the dyno session proved it was insufficient at this level of power. The tube and fin IC taken from the Lancer Evo was good up to 350-400 crank hp. I did choose to go the tube and fin way (and not the bar and plate) for a simple reason: Greece is a country with very hot weather in general, so i needed to take advantage of the rapid cooling effect of tube and fin. Think about it in this way: bar and plate has much more material on it, it is heavier as well, which means that if it gets hot, its efficiency drops a lot, since it is very difficult to cool down again. If the same car was used in Sweden where temps are freezing, then a bar and plate IC would be more efficient, since it would hold low temps in it much more efficiently than a tube and fin. The size i chose to go with was a core of 500x400x50mm. With the side panels it fitted exactly between my lights panel, but the height required a bit of fabrication to fit. It was decided to go with a taller and thinner size since height gives much more cooling capacity than thickness. Here is an interesting read from Garrett comparing horsepower capabilities vs sizing http://www.turbobygarrett.com/turbob...t/intercoolers















    Another thing that we decided to do is eliminate the problem that appears on tube and fin IC's (pressure drop) and weld some inverted V shaped pieces of metal between the tubes. This would (in theory) help air not to get stacked up between the tubes which increases pressure drop.



  • The piping and TB needed to be larger. As it proved the piping and TB should be large enough not to increase air velocity over a limit. For a given air flow, in my case about 50-60 lbs/min, the pipes should allow air to travel at about 200-300 ft/sec. According to my calculations i decided to go with 65mm ID tubing which gives a bit over the limit air speed. Here is an interesting read by Garrett for this issue https://turbobygarrett.com/turbobyga...ge_air_cooling
    Following the piping, the throttle body should become bigger as well. The stock was 54mm, which is really small for this kind of power. First thoughts were to bore the stock TB, but this would take me safe up to 58mm. It was decided to fit a TB from a BMW, 65mm on the stock intake. Hopefully the intake will not be much of a restriction now.











  • Relocated oil cooler, since the intercooler took its original place. The oil pipes were shortened and it was fitted with rubber mounts so that nothing breaks. To make sure that it will stay on the car even if the rubber mounts break, the mounts were secured with a piece of square band, so in the case of failure the bands will hold the cooler in place.









  • Had a sprung custom 6 puck clutch disk made. The stock 850 organic clutch disk for some reason held about 530NM of wheel torque combined with the racing Sachs pressure plate, while nobody was expecting it. However these are big numbers for a 228mm clutch disk AND organic materials. I am expecting a raise of torque, so i figured that i should do anything i can to hold the power as good as possible at 228mm. In any way i am close to M90's power capabilities so there would be no point making a new flywheel + clutch disk + pressure plate for it. If it breaks i will change over to a 420G BMW transmission and swap all of these things. One more thing that seemed logical to do was to buy a spigot shaft bearing adapter from RetroTurbo, so that the shaft remains as rigid as possible to protect it from breaking.



  • Replaced the 8 stock 12.9 flywheel bolts with ARP. Very expensive stuff from KLracing, however i wanted to have maximum protection from a flying flywheel.

  • New turbo GTX3076R plus twin scroll 0.83 turbine housing. Since the GTX2867R was maxed out, the next option was a step up the Garrett range of turbos. The 3076R gives the capability of 640 hp, so it should cover my 500whp target. It was combined with a 0.83 TS housing, since it is said that this kind of housing spools like it was the 60% of its size and performs maximum as the result of the its size divided by 1.20. This means that it would spool like 0.50 A/R and its performance is like 0.7 A/R. These are the theoretical values by Garrett. Of course this meant a bit of manifold porting, and downpipe modification. The new turbo is actually shorter than the older setup. It is impressive though how much bigger is the 3076R's wheel compared to the 2867R's. Also changed over to a PTP turbo blanket. Both of my 70 euros blankets that i bought off ebay were burnt and torn both being titanium lava etc. Be advised .. don't get cheap stuff for this part !









  • New cone filter by K&N. Despite the fact that i wanted to go for a larger cone filter, the ones that flow more than 950CFM are absolutely HUGE !! For anyone interested the 4 inch diameter cone filters that i wanted to choose from were the following:

    RD-1460 = 1187CFM
    RU-5045 = 1130CFM
    RE-0870 = 900CFM
    E-1997 = 873CFM

    I got the values from K&N reps. Also found out that inverted cone filters may cause problems to MAFs and their flow can not be calculated. So you better avoid them. I went for the RE-0870, which as you can see marginally fitted in my setup. Unfortunately i will have no cool air feed from anywhere, but there is nothing more i can do. Maybe removing my headlight ?



  • Switched the thermostat to a 82*C. Thought that keeping temps a bit lower would be good for the engine, since it is pushed so much. After runs the temps were about 92*C so now this will probably be improved.

Here is how the engine bay's appearance is now, hopefully it won't need to be changed again..



For now i will hold my breath up to when the weather reaches 10*C to have an advantage for the dyno session. And .. cross your fingers that my transmission holds ...

PS: If you want to see more detailed pictures you can visit my Photobucket HERE. Make sure you check all the sub-folders. However Photobucket is crap. It messes with my photo sub-titles, can not save custom pics arrangement. Sorry about this !!




UPDATE 24/04/2017

After a rough winter in Greece (in terms of politics not weather), i finally have some things to present to the forum !! Despite being late due to several reasons, and while the weather is getting warm again, the -almost last- dyno session results are below. Some observations and conclusions for the ones who are interested in this project:
  • The intercooler theory proved to be very "correct". Intake temps are at maximum 8-9 degrees C over the ambient temperature in the dyno, while on the road it usually tops to 5*C difference.

  • On dyno sheet boost does not represent real situation. The dyno operator keeps his foot on the gas at a constant 2500 to 3000 rpm, and then floors it, probably to protect transmission. This means that boost starts to build from this rpm. However on the road if for example i am on 4th gear and floor it while rolling at 1500 rpm, i see 1 bar of boost at 2800 rpm which of course tops at 2 bar before 3500 rpm. GTX3076R is the absolute perfect (big) turbo for this car. Seems to perform much better than the GTX2867R but remember that the 3076R has a twin scroll hotside.. Fully recommended if you build a TS manifold !!

  • It seems that the limits of this head are close. This may not be a very accurate conclusion, since the limits could rely on the intake manifold, cam, valves or seats, but for sure something in this area, which to investigate would be VERY expensive from this point on. Could be addressed with extra boost but there is also a 100 octane unleaded fuel limit. The reason i am saying this is that the engine when pushing the timing to produce more power is much more sensitive to knocking compared to the previous turbo, where the timing could be stretched to insane values. Of course this time the air volume and pressure entering the cylinders is much greater. Another thing to observe is that moving the cam gear from 0 degrees (pull no 30) to a few advanced degrees (pull no 31) and then to some retarded (many - 6 or 7) degrees (pull no 32) does not do much difference on whp neither moves the curves around as usual. Only gain was that on very high rpm the power remains flatter. Pull 32 is better than 30 which is better than 31. Also there is no point taking the engine much higher than 6500-7000 rpm as it seems.

  • Freeze plugs have to be secured in absolutely every engine. I did not believe that they could pop, noone here had a similar previous experience, but they did during the dyno session, which stopped us from investigating the cam gear / timing puzzle. Mine were the old style 45 mm dome style since it is a B21, which were replaced when i overhauled the engine. The second cylinder's plug though changed its mind and wanted to escape, and so it did !! I installed new ones, and made some TB style brackets to hold them, while drilled and tapped the block at some reinforced points. Unfortunately the ready solutions were for the dished style freeze plugs, so i had to improvise. Hopefully issue is solved now. I think this is due to harmonics vibrations of the block or some freeze plug defect, but i would not like to see them pop again ...






  • Apart from the cam timing i will try some proven good gasoline additives to achieve knock resistance, raising the octane number and try to play with engine timing / boost, to see if the power will be the same. Have to note though in one pull that we stopped due to hard knocking we saw 400whp at 4000 rpm before stopping. Sounds promising for racing or other type fuels for this kind of engine! Note that in this state to keep an AFR of 12 under full boost, 85% of the ID1000s @ 3 bar is used, ie 3400cc/min = 3.4 lt/min, which means that it consumes the tank of 60 lt in 17 minutes if it could be used in full throttle for that much .. Math is impressive sometimes :P

  • Turbo sounds are awesome with this cone filter and turbo







Other tasks performed:
  • Replaced odo gear since the old one failed, lost 2 teeth. Fortunately after good cleaning and oiling the gears, worked perfectly at once !

  • Cluster Voltage Regulator. Since i had the cluster out of the car, thought of measuring the voltage regulator that served the car for many years. Found out that it stabilized the voltage correctly only when input was up to 11.90 Volts. Above that limit it gave as output more than 10 Volts .. ie between 11 and 12 when the input was between 13,5 - 14,5. So i replaced it with a 7810. Take care though that the 7810 has to be mounted upside down since its pins are inverted compared to the old voltage regulator. Only thing required is bending the 7810's pins a bit and some grinding of the metal bracket that mounts on.






UPDATE 12/02/2018

After a long time of inactivity, or more precisely waiting for a shop to complete "dressing up" my new seats for the front and my old ones for the rear, it was time to install them ! Along with the seats i had a few more interior changes, so to look better than before. Also a minor engine bay improvement.

So the list of changes this time is:
  • The front new seats were -surprisingly for the Greek market- taken from a Lancia Delta Integrale. They are genuine Recaro, however since they are very old needed lots of work to restore them. Other than that, i kept the original Volvo railways, which very easily adapted to the new seats, using as and adapter a thick 90 degrees bent metal sheet. On driver's side i had to cut out the tilting functions that the original seat had, but by drilling a few holes on the adapter the seat angles can be achieved by screwing them into different slots if needed. The plastic parts of the seats bolted on with no mods. On the co-driver's seat was also integrated a seat reminder sensor taken from an S60 (part no 31413337), and works nicely with the stock circuitry despite being totally different .. The rear seats remained the same, but were also restored with the same fabric as the front.









  • Made a new custom shift stick cover




  • Put a new carpet on the rear deck and altered my round ancient Pioneer speakers to 6x9 by Sony, a company i don't like, but they have seriously good sound for my needs. I am thinking of altering my front speakers as well. Also ordered an Alpine KTP445U amplifier to install stealthily behind the dash somewhere ! Elvis says hello



  • Replaced my old leaky turbo circulator which was part no 0392020016 (discontinued) with a new by Bosch 0392020024. Despite having different outer input/output dimensions (old = about 15mm - new 20mm), the inner hole is 1-2 mm different, so i assume the flow/pressure numbers are the same. But in any way i just need it for circulating the water after the engine is off. Only problem was that i had to make new adapters so to fit on my setup. Their power supply plugs are also different.

    Since i hit the limit for pics for this post, check the next story post for pics of the circulators.

Continued at post #82 --->LINK<---

Last edited by aristg; 04-02-2018 at 03:56 AM.. Reason: f*$#k photobucket for 3rd time
aristg is offline   Reply With Quote