[MileHighScottie]

ArvidW, I didn't see it in any photos but did you fabricate a heatshield between the turbo and the engine? I love how tidy this set up is.

 

[ArvidW]

Correct, there is no heatshield between turbo and engine. I think there will be sufficient airflow around the turbo to prevent the powder coat melting on the crankcase.

 

[ArvidW]

Today I tested and added the wastegate actuator to the bike. A custom bracket had to be build because I switched it to the hotside due to clearance issues on the cold side.

The wastegate actuator starts moving at 0.3bar and reaches it full 12mm stroke at 0.7bar. I'll set it to 0.4bar wastegate cracking pressure (3mm preload) and increase it on the dyno if neccessary (I'll be using a bleed solenoid controlled by the ECU as well)

Clearance to the oil cooler is limited, but sufficient:

 

[ArvidW]

I started working on the intake runner parts today. First experimented a bit to get the intake runner radius nice and smooth (5 different angles from 0-90dg, afterwards smoothed with a file & sandpaper). Later started working on three identical stacks for the intake runners. All runners have a lip to prevent the hose from slipping off:

I decided to build something different than the usual boxed intake with short straight runners. I'm planning on building a narrow trapezium shaped intake plenum with long intake runners. The long runners will most likely rob some power due to incorrect tuned length. But I hope to compensate that by having a better & more even distributed flow through the plenum. Furthermore access to the hose clamps etc will be better with this solution.

Edit: According to http://www.exx.se/techinfo/runners/runners.html link the tuned runner lengths for first and second harmonics are:
- Stock (aprox 22mm to valve) = 8200 & 16400rpm
- Modified (aprox 36mm to valve) = 5000 & 10000rpm

The new runner length should fatten up the midrange to some extend. And If I add 3cm more length I may even use the second harmonic at redline..

 

[ArvidW]

Airbox under construction:

 

[ArvidW]

The airbox is finished. Idle valve and boost reference point are build in and properly sealed:

Also fabricated the last intake hose including the original IAT sensor:

But the best thing is: It runs, AND makes boost !

The bike is now running boost with the stock ECU (i know, Turbo and Alpha-N without MAP multiplication is a terrible idea). I only modified the >80TPS fuel and ignition tables so it can only "safely" run boost at WOT, which is good enough for the first testdrive.

It's quite a bit quicker than the stock bike, but with the limited boost and "safe timing and fat fuelling" it ain't the rocket ship it can be. But the clutch was already severely slipping at 4500rpm.. (fixed it temporarily by increasing preload, and removing belleville washer)

The feeling is wonderful though. It builds boosts in 0.5-1second, already at 3500-4000rpm. And just pulls and pulls, until you close the throttle and you get rewarded by the sound of the compressor surging. The bike is so much more quiet with the closed airbox, sound damping of the intercooler and turbine in the exhaust.


Next job: engine management so it can actually start making some decent power!

 

[ArvidW]

A MAP sensor has been added to the Arduino datalogger so I can finally measure the boost:
- 0.3bar @4800rpm, 0.22bar @9500rpm (wastegate pressure, with little preload & no boost controller yet... )
- 1.7s between WOT (3225rpm) and full boost (4863rpm), which is quite slow but the start RPM was very low as well.
- Only 6dgC increase in intake air temperature during 8s WOT run (open type sensor)
- AFR is a still bit of a mess, so i'm glad it only makes 0.3bar at the moment!
- Clutch slips from 5000rpm and engages at 5800rpm (see kink in rpm signal)

If you go WOT from a higher start point in a lower gear (5241rpm) boost builds a lot quicker at 0.7seconds:

Very pleased with these first results, it's nice to be able to test the bike at a low boost level. The clutch is crap though..

 

[MileHighScottie]

Wonder what will solve the clutch slip. I forget if you tried stiffer springs yet?

Question about the AFR's. If I'm reading it right, you're pretty rich under boost, down in the mid 10s. Better than lean but are you going to tune those close to Stoich?

I cant wait till this thing is on a dyno

 

[ArvidW]

Wonder what will solve the clutch slip. I forget if you tried stiffer springs yet?

Question about the AFR's. If I'm reading it right, you're pretty rich under boost, down in the mid 10s. Better than lean but are you going to tune those close to Stoich?

I cant wait till this thing is on a dyno

The clutch already has:
- EBC 10% stiffer clutch springs (since 2015)
- 2mm more preload on clutch springs (shims)
- Removal of anti-judder spring (it applies pressure to last seperator plate, reducing clamping load on last friction plate of the clutch pack)
- Clutch lever extension @ engine (Triumph option if i remember correctly)

But I suspect the clutch has been bad from the moment I've bought the bike (5600km old bike). The clutch didn't have a positive engagement when launching hard even when the bike was stock, so I bought the EBC springs to fix it.

Time to replace the clutch plates? (bike is now 25000km old)

 

[DaveM]

This is a most educational and interesting thread ArvidW.
Clearly you are a talented fellow, with many skills.

Cheers
DaveM

 

[ArvidW]

@DaveM, thanks a lot! I've alway's appreciated other people sharing their projects so i'm happy to share as well

@MileHighScottie, This weekend I've measured the clutch plate thickness:
- Friction plates: 3.974mm on average (new 3.80-3.88mm, service limit 3.60mm)
- Seperator plates: 1.994mm on average (new 2.00mm, no service limit)

So the plate thickness is OK.

I've consulted some colleagues responsible for the wet clutch development (Dual Clutch Transmission) and their professional verdict is that the friction material has been damaged due to high temperatures and slip speed :whistle .. So i guess i'll be buying new ones soon.

Does anybody know if the ST1050 friction plates and separators fit the 955i engine? They look the same, but the ST1050 has thinner and 1 more friction plate... That would already give me 11% more torque capacity.

The 955i clutch can theoretically support:
MengMAX = (mu*nf*ns*fs*Reffect)/i = 124Nm
mu = 0.16 (most wet friction plates are between 0.14 and 0.18)
nf = 9*2 (number of friction sides of friction plates)
ns = 5 (number of springs)
fs = 230N (preload of spring on clutch, calculated from EBC spring dimensions)
i = 105/60 (primary drive from clutch to crank)
Ro = 139.37/2mm (outer radius friction plate surface)
Ri = 121.23/2mm(outer radius friction plate surface)
Reffect = 2/3*(Ro^3-Ri^3)/(Ro^2-Ri^2) = 65.3mm (130.5mm diameter)


So I'll need to improve it >20% to be able to hold 150Nm... :|

 

[ArvidW]

Good news: A ST1050 clutch fits the 955i engine!
Even better: The clutch doesn't slip anymore (at this boost level...)

Clutch friction & seperator plate splines and overal size is identical. The main difference is that the ST1050 friction plates have a larger inner diameter and thinner plates creating room for the 10th friction plate.

Stack consists of:
2x outer friction plate (3.8mm, same as 955i)
8x center friction plate (3.3mm, 0.5mm thinner than 955i)
4x seperator plate (2.0mm, same as 955i)
5x seperator plate (1.6mm, 0.4mm thinner than 955i)

Total stack hight is 50.2mm compared to 50.1mm from the 955i pack (previous post states friction plate thickness incorrect for the 955i, it should be 3.794)

I think it is possible to make a 11plate clutchpack. If I take 11x 3.3mm friction plates + 10x 1.6mm separators the pack will be 52.3mm, which fits within the basket. The anti-judder spring doesn't fit in the pack anymore, but in my opinion that's acceptable This may be an option for the future if the clutch starts slipping again.

 

[ArvidW]

During Black Friday a local speedshop (jt-performance.nl) gave a nice discount on one of their standalone ECU's:

So i'll be running a EMU black on the RS . It's a very nice ECU as it has loads of I/O, inbuilt wideband controller (also for the newer LSU4.9 sensor!) and is IP67 water resistant.

 

[ArvidW]

It's been a while, some progress has been made:

A knock and cam sensor have been added, including a lot of wiring.

And changed the trigger setup from "missing & additional tooth" to "multitooth" as the ECU didn't support the original setup.

 

[ArvidW]

Bike runs great with the new ECU. Lambda is now stable and way easier to control

It now runs 0.25bar of boost, which is already a very nice ride. I'll crank it up to 0.55bar but only up to 7000rpm as the injectors will be maxed out. :crying: The boost will be ramped down to 0.2bar @ 9600rpm so I can keep the injectors at 90% dutycycle for now

 

[MileHighScottie]

Wow she's beautiful in that yellow.

Can you get bigger injectors so you don't run out above 7000rpm?

I'm looking forward to seeing this thing on a dyno

 

[HAP]

I have long had a love of Triumph turbo kits but cost and dependability have prevented me from buying a kit. Lack of skilled prevents me from creating my own

The potential of attaching a kit to a well loved and cared for 1999 Daytona is still very much a "want" item and following the great thread and the exceptional work you have done on your RS gives hope that I may yet get the power and dependability of a new, off the line, litter bike from a 19 year old, super sorted Daytona that fits me like my own skin.

When you get around to the completion stage of yours, please give thought to creating one for others to enjoy.

John King had a 2002 Turbo Daytona that left me on his 999r in the dust. Yes, he probably is a better rider than I but I have been hooked on that idea that a bolt on kit would be an up grade still cheaper than a hyper bike costing thousands more.
Thanks for sharing and the work is amazing. Can't wait till you pronounce it complete.
Well done

 

[ArvidW]

@MileHighScottie, bigger injectors would indeed be a logical choice, unfortunately it's 200euro that's currently not in the budget. I did however find a cheap 4bar pressure regulator that increases the injector flow by aprox 15%.

This should lift the 90% dutycycle point to aprox 8200rpm @ 0.55bar of boost with lambda 0.8 (AFR 11.8). Above 8200 i'll drop the boost off to not exceed the 90% injector dutycycle.

It won't yield the highest HP number on the dyno (aprox 145 at the crank) but the torque will be equivalent to the "turboconnections 955i" kit from 2000, with very similar wheely potential.

@HAP, Thanks for the compliments, much appreciated Most kits where around 5000dollars back in de day I suppose? Until now i've invested 1800euro, which includes the ECU (850), turbo (100) and primarily small items like tubing(200), wiring(50), connectors, sensors(100) and couplers. There's 200euro left for dyno-tuning until I exceed the purchase price of the bike.

 

[sussurf]

Arvid did you consider blueprinting the TB's or is that redundant in forced induction?
Love your work mate!

 

[ArvidW]

@sussurf, Blueprinting to increase airflow is certainly not redundant on a turbo engine. Most things you can do to improve airflow through an NA engine will also be beneficial in forced applications

I didn't spend time on it though. Building the entire turbo setup from scratch was already sufficient work to keep me busy.