AWD Civic Coupe Turbo Build
Discussion
Thanks mate - i didnt want to post a link personally in case i ended up getting shot for it by the admins.. 
Hopefully the engine will sell, it hasnt this morning but there are a few people watching it so hopefully with a bit of time.. plus who can afford that sum of money out of the blue (wishful thinking!)
I cannot do anything until the engine sells, which will in turn fund most of the additional work including new engine.
Hopefully the engine will sell, it hasnt this morning but there are a few people watching it so hopefully with a bit of time.. plus who can afford that sum of money out of the blue (wishful thinking!)
I cannot do anything until the engine sells, which will in turn fund most of the additional work including new engine.
Hi there,
its been a while since ive made any progress on this build - mainly due to work commitments.
However i do have some progress to report. This is my new engine;
Its a W8 4.0 taken from a passat, ive been working on a brief rebuild and checked the engine for any damage / wear. I am just finishing up rebuilding it with the intent to put two turbochargers on it, for now i wont be decreasing the compression ratio but ill be designing a head gasket spacer to drop it to a sensible level. There are many reasons to be careful with this engine, its benefits of being compact also means thin connecting rods etc.
I have decided to move way from the transmission / engine combo and use this one. The audi 01e transmission is perfect and significantly stronger than the crv model. I have also managed to source a honda rdx rear differential to replace my crv differential which allows me to experiment a bit with torque vectoring
its been a while since ive made any progress on this build - mainly due to work commitments.
However i do have some progress to report. This is my new engine;
Its a W8 4.0 taken from a passat, ive been working on a brief rebuild and checked the engine for any damage / wear. I am just finishing up rebuilding it with the intent to put two turbochargers on it, for now i wont be decreasing the compression ratio but ill be designing a head gasket spacer to drop it to a sensible level. There are many reasons to be careful with this engine, its benefits of being compact also means thin connecting rods etc.
I have decided to move way from the transmission / engine combo and use this one. The audi 01e transmission is perfect and significantly stronger than the crv model. I have also managed to source a honda rdx rear differential to replace my crv differential which allows me to experiment a bit with torque vectoring
Hey all,
thanks for the comments, not the first time my sanity has been questioned - oh well im nothing if not eccentric..
I do look forward to the phone call with the insurance company, i think i might need to buy shares of some kind!!
I rather rushed me previous post as i didnt think people would be too interested, after all its more of an engineering project to me as opposed to a car :S
There are a whole raft of jobs which needs to be done, specifically;
Custom exhaust manifolds- not an issue however i do need to find someone who can machine me two flanges based on the exhaust gaskets as guides.. Then ill fabricate the stainless SCH10 tubing. Ive designed twinscroll exhaust manifolds so i will be aiming to still exploit the GT3076r and two of them..
Fabricate inlet manifold- need to blank off the single throttle body, procure a second throttle body and using a pair of custom ali flanges and ali tube i am going to add both throttles to the sides of the inlet manifold. One per turbo.
Engine management - ive scrapped the OEM VW ecu in favour of an adaptronic 1280s - which in my opinion is an absolutely staggering ecu.. FPGA based it is capable of almost anything. I have since programmed it with dual drive by wire throttles and also used my anti lag system from the honda engine. So this 4.0 W8 will still use the advanced, rally style anti lag system to reduce lag. I have also designed an algorithm and system which will control the wastegate dutycycle based on the rotational speed of the turbos, ensuring that although the net output boost control will be maintained the amount of bypass will be varied to ensure that both turbos shaft speeds stay similar - this should reduce the effect of one turbo slowing more than the other and having to recover that rotational inertia when the throttle is reopened..
Transfer case - since i am using the O1E transmission, the (was) front differential will now run the rear wheels - which i hope to upgrade at some point to a Quaife ATB. The (was) rear wheel output i am going to use a silent chain drive to offset it and feed it forwards, past the drivers side of the engine to the original Honda Prop shaft. This will then travel diagonally down the car to the modified Acura RDX differential in the front. The benefit here is that there is no actual differential - the amount of torque imparted to the wheels is controlled by an electronic clutch and planetary gearset. This will allow me to control the front to rear torque transfer as well as right to left. I will be integrating this with throttle input, steering angle etc. to give me torque vectoring. This should help pull the car through turns without the dreaded awd understeer issues. Since this is all my design i will also be able to change the control methods at any time and refine them over the coming years.. which is pretty nifty
thanks for the comments, not the first time my sanity has been questioned - oh well im nothing if not eccentric..
I do look forward to the phone call with the insurance company, i think i might need to buy shares of some kind!!
I rather rushed me previous post as i didnt think people would be too interested, after all its more of an engineering project to me as opposed to a car :S
There are a whole raft of jobs which needs to be done, specifically;
Custom exhaust manifolds- not an issue however i do need to find someone who can machine me two flanges based on the exhaust gaskets as guides.. Then ill fabricate the stainless SCH10 tubing. Ive designed twinscroll exhaust manifolds so i will be aiming to still exploit the GT3076r and two of them..
Fabricate inlet manifold- need to blank off the single throttle body, procure a second throttle body and using a pair of custom ali flanges and ali tube i am going to add both throttles to the sides of the inlet manifold. One per turbo.
Engine management - ive scrapped the OEM VW ecu in favour of an adaptronic 1280s - which in my opinion is an absolutely staggering ecu.. FPGA based it is capable of almost anything. I have since programmed it with dual drive by wire throttles and also used my anti lag system from the honda engine. So this 4.0 W8 will still use the advanced, rally style anti lag system to reduce lag. I have also designed an algorithm and system which will control the wastegate dutycycle based on the rotational speed of the turbos, ensuring that although the net output boost control will be maintained the amount of bypass will be varied to ensure that both turbos shaft speeds stay similar - this should reduce the effect of one turbo slowing more than the other and having to recover that rotational inertia when the throttle is reopened..
Transfer case - since i am using the O1E transmission, the (was) front differential will now run the rear wheels - which i hope to upgrade at some point to a Quaife ATB. The (was) rear wheel output i am going to use a silent chain drive to offset it and feed it forwards, past the drivers side of the engine to the original Honda Prop shaft. This will then travel diagonally down the car to the modified Acura RDX differential in the front. The benefit here is that there is no actual differential - the amount of torque imparted to the wheels is controlled by an electronic clutch and planetary gearset. This will allow me to control the front to rear torque transfer as well as right to left. I will be integrating this with throttle input, steering angle etc. to give me torque vectoring. This should help pull the car through turns without the dreaded awd understeer issues. Since this is all my design i will also be able to change the control methods at any time and refine them over the coming years.. which is pretty nifty
I hadnt realised there was a small update i missed a while back - quite a while back - hence the surface rust here..
I opted to replace the electro-hydraulic steering system i was using previously with this, its an electric power steering rack from an FN2 Civic Type R.. The logic being that the hydraulic system was a bit rubbish and this system seemingly the new standard uses a combination of electric motor and strain gages to determine the driver demand and reduce it using the motor.
What i have done is intercept the signals from the strain gages and run them into a ADC, DAC, this allows me to intercept and modify the signals. With this i now have a completely tunable power steering system which i can modify the "weight" of at any moment. The first job when it hits the road is to override the OEM map by providing an inverse signal which will give a flat response to driver torque over speed. Once this is done i will then be able to run my own map on top based on whatever sensors or parameters i want. This is a safety critical system so i also have a bypass switch to disconnect the interface and run it as normal once again.
Its given me a fair amount of control over the system without spending alot of money whatsoever. The only issue is that it isnt 100% physically compatible so i needed to have some adapter parts made.. Firstly it sits on a modified subframe - please ignore the surface rust, most of this is being scrapped for the new differential. To connect the rack to the hubs i had some spacers made from steel (safety sake over aluminium) and a set of rose joints on the end. The travel of the rack is actually slightly more than the original too, which will benefit the maximum steering lock- i cant quantify how much of a benefit this is at the moment though..
When the new differential arrives i shall be finishing up evertyhing in the engine bay and tidying it up - alot...
I opted to replace the electro-hydraulic steering system i was using previously with this, its an electric power steering rack from an FN2 Civic Type R.. The logic being that the hydraulic system was a bit rubbish and this system seemingly the new standard uses a combination of electric motor and strain gages to determine the driver demand and reduce it using the motor.
What i have done is intercept the signals from the strain gages and run them into a ADC, DAC, this allows me to intercept and modify the signals. With this i now have a completely tunable power steering system which i can modify the "weight" of at any moment. The first job when it hits the road is to override the OEM map by providing an inverse signal which will give a flat response to driver torque over speed. Once this is done i will then be able to run my own map on top based on whatever sensors or parameters i want. This is a safety critical system so i also have a bypass switch to disconnect the interface and run it as normal once again.
Its given me a fair amount of control over the system without spending alot of money whatsoever. The only issue is that it isnt 100% physically compatible so i needed to have some adapter parts made.. Firstly it sits on a modified subframe - please ignore the surface rust, most of this is being scrapped for the new differential. To connect the rack to the hubs i had some spacers made from steel (safety sake over aluminium) and a set of rose joints on the end. The travel of the rack is actually slightly more than the original too, which will benefit the maximum steering lock- i cant quantify how much of a benefit this is at the moment though..
When the new differential arrives i shall be finishing up evertyhing in the engine bay and tidying it up - alot...
purplecivicturbo said:
Custom exhaust manifolds- not an issue however i do need to find someone who can machine me two flanges based on the exhaust gaskets as guides..
Take std gasket, place on scanner, scan 1:1. Import .BMEP into freebee cad package (fastcad,sketchup etc), draw 0mm width lines over outside perimiter/holes in gasket. Export file in .dxf format. Email to one of hundreds of water jet cutters, wait a few days, new flanges arrive in post! Simples and cheap!What isn't either of those things is:
purplecivicturbo said:
control the wastegate dutycycle based on the rotational speed of the turbos
To do this you have to measure the speed of the two turbos........ (based on doing this for years in both wrc and sportscar racing, i can tell you, do not underestimate the task to reliably measure the shaft speed of a turbocharger, let alone two!)hey mate, thanks for the comments.
I am glad we agree - although i hadnt thought about waterjet that is the exact method i was hoping to use - i kept my old gaskets to scan in at work whilst nobody is looking..
With regard to the turbo speed issue, i also agree, reliability is an issue. Depending on the material and the sensing method. Ive been doing work with VR sensors and hall sensors trying to get it bang on and reliable. Borgwarner's EFR turbos come with speed sensors, so i am hoping to use something like that.
Plus you can use a kalman type filter to predict the turbo speed based on previous samples. This will effectively ignore sporadic results and yield a quick response to change - although i am still yet to experiment with that i completely agree that it is a technical hurdle to overcome. In terms of computing the duty cycle changes i will be using a fixed positive pressure source in order to mitigate some of the changing variables and use fuzzy logic (or something similar) to compute the duty cycle changes based on demand.
As i say this is more of an engineering challenge to me than anything else, if i cannot manage to do it itll be a loss but i just want to rise to the challenge since i think the results will be worth it.. in the end... hopefully... with any luck...
I am glad we agree - although i hadnt thought about waterjet that is the exact method i was hoping to use - i kept my old gaskets to scan in at work whilst nobody is looking..
With regard to the turbo speed issue, i also agree, reliability is an issue. Depending on the material and the sensing method. Ive been doing work with VR sensors and hall sensors trying to get it bang on and reliable. Borgwarner's EFR turbos come with speed sensors, so i am hoping to use something like that.
Plus you can use a kalman type filter to predict the turbo speed based on previous samples. This will effectively ignore sporadic results and yield a quick response to change - although i am still yet to experiment with that i completely agree that it is a technical hurdle to overcome. In terms of computing the duty cycle changes i will be using a fixed positive pressure source in order to mitigate some of the changing variables and use fuzzy logic (or something similar) to compute the duty cycle changes based on demand.
As i say this is more of an engineering challenge to me than anything else, if i cannot manage to do it itll be a loss but i just want to rise to the challenge since i think the results will be worth it.. in the end... hopefully... with any luck...
Gassing Station | Readers' Cars | Top of Page | What's New | My Stuff