Power Losses when retarding Ignition
Discussion
Hi all, my name is Lee! This question may sound a bit odd, but what kind of power losses will I see by retarding the ignition and what is a safe amount?
Im involved in a university project called formula student where universities from around the world build and race single seater open wheeled racecars with a 600cc motorbike engine.
The reason I ask is im an Electronic Engineer and my project is a traction control and launch control system. One of the ways I wish to stop the wheel spin is retard/cut the spark.
The engine will probably be a supercharged 550cc aprilla V Twin, with about 80bhp @ 12000 RPM.
regards Lee S
Im involved in a university project called formula student where universities from around the world build and race single seater open wheeled racecars with a 600cc motorbike engine.
The reason I ask is im an Electronic Engineer and my project is a traction control and launch control system. One of the ways I wish to stop the wheel spin is retard/cut the spark.
The engine will probably be a supercharged 550cc aprilla V Twin, with about 80bhp @ 12000 RPM.
regards Lee S
Have a look at megasquirt, indeed the megasquirt-extra variant that has different stratergies for spark/fuel cutting. They have a crude version of TC called anti-rev, which just looks at the rate of change of RPM and cant limit revs.
For rev limit, on my v8 i retard ignition to 5 degrees, which seems to work well - it has hard fuel cut at 5700 rpm but has only hit that once.
For rev limit, on my v8 i retard ignition to 5 degrees, which seems to work well - it has hard fuel cut at 5700 rpm but has only hit that once.
Retarding the ignition a significant amount will increase tempretures in the exhaust, which may affect exhaust valve and sead longevity and my provide a source of pre ignition.
Cutting the ignition will give you unburnt fuel in the exhaust, which is less of a problem unless you run a cat.
Steve
Cutting the ignition will give you unburnt fuel in the exhaust, which is less of a problem unless you run a cat.
Steve
There's no rule of thumb.
It depends on the engine and how steep the "ignition loop" is.
An ignition loop is a plot of data on a graph with measured torque along the Y axis and ignition advance along the X-axis.
This will produce a curve- where the gradient = 0 or dy/dx = 0 is the point of best torque or MBT. If the ignition loop is steep you'll see alot of change in torque for a small change in ignition.
The ignition loop will be steeper for full load/WOT than at part load.
The ignition loop will be steeper at peak torque WOT than at nearer 1000 rpm WOT.
An ignition loop will be significantly steeper for a boosted engine than a naturally aspircated engine.
An engine with a high CR will tend to have a steep ignition loop at WOT.
An engine that breathes well with high volumetric efficiency will tend to have a steep ignition loop.
Those are some of the variables that effect this phenomemon
It depends on the engine and how steep the "ignition loop" is.
An ignition loop is a plot of data on a graph with measured torque along the Y axis and ignition advance along the X-axis.
This will produce a curve- where the gradient = 0 or dy/dx = 0 is the point of best torque or MBT. If the ignition loop is steep you'll see alot of change in torque for a small change in ignition.
The ignition loop will be steeper for full load/WOT than at part load.
The ignition loop will be steeper at peak torque WOT than at nearer 1000 rpm WOT.
An ignition loop will be significantly steeper for a boosted engine than a naturally aspircated engine.
An engine with a high CR will tend to have a steep ignition loop at WOT.
An engine that breathes well with high volumetric efficiency will tend to have a steep ignition loop.
Those are some of the variables that effect this phenomemon
Eliot, we get a majority of points from costing and I have been looking at using Megasquirt for the fueling and spark.Im not sure how much of a hassle this would be instead of using an aftermarket ECU such as Motech? However i must design the traction control system as part of my final year project.
Steve thanks for the info, hopefully the traction control wont be needed too much, so we will only be retarding the ignition 2-3 times a lap, for maybe a second or so. The cars run on slicks in the dry. The retard and cut will depend upon slip.
Marquis, again thanks for the useful info. We have a dyno at university so maybe getting the ignition loop could be something the engine guys could do. From what you have told me though it sounds like the ignition loop could have a high dy/dx.
If this proves too hard I will cut the spark to alternate cylinders to stop the plugs fouling. We arent running a cat so I think this should be o.k.
regards Lee S
Steve thanks for the info, hopefully the traction control wont be needed too much, so we will only be retarding the ignition 2-3 times a lap, for maybe a second or so. The cars run on slicks in the dry. The retard and cut will depend upon slip.
Marquis, again thanks for the useful info. We have a dyno at university so maybe getting the ignition loop could be something the engine guys could do. From what you have told me though it sounds like the ignition loop could have a high dy/dx.
If this proves too hard I will cut the spark to alternate cylinders to stop the plugs fouling. We arent running a cat so I think this should be o.k.
regards Lee S
I think traction control systems like this are best treated as giving the driver a 'window of stability' to play in, rather than taking ultimate control and letting you bury the right foot in any gear under any conditions. If you use it like that, you don't need to be able to take masses of power away for very long. In any case it's usefull for the driver to be able to 'drive through' the TC and provoke a significant amount of slip under some circumstances, so you probably don't *want* the TC to have too much control authority. What is useful though is to make it obvious to the driver when the TC is engaging - either via a big light, or by introducing a perceptable stutter. Just backing the ignition off wouldn't give any obvious feedback to the driver. A walking spark cut would be more effective from that point of view - as well as kinder to the valves/exhaust.
Green thanks. I dont think any of our drivers will have any race experience so the TC will provide a significant safety net. My first car was a mk1 Escort so I know the consequences of a heavy right foot! I dont expert people will be burying the throttle all the time but it will come into play more than it would with a seasoned racer. It will probably be a good advantage if it rains. I plan to have variable slip depending upon the conditions and if the car is in a turn or not. A warning light is a good idea!
Edited by LeeSpeakman on Friday 29th September 14:34
Eliot, that sounds like a great idea, and its something id look into for one of my own projects as ive been looking at megasquirt for a turbo xflow im collecting bits for.
Problem is the TC system will count to 25% of my Degree and I dont think modifying existing software will be what they intend of me.
Am i right in thinking that Megasquirt has additional input / output ports? If so then I might be able to interface the Traction Control Logic to the ECU instead of having to interupt / delay the spark with another circuit - just another link in the chain so to speak ......
My supervisor might go for that!
Problem is the TC system will count to 25% of my Degree and I dont think modifying existing software will be what they intend of me.
Am i right in thinking that Megasquirt has additional input / output ports? If so then I might be able to interface the Traction Control Logic to the ECU instead of having to interupt / delay the spark with another circuit - just another link in the chain so to speak ......
My supervisor might go for that!
LeeSpeakman said:
Problem is the TC system will count to 25% of my Degree and I dont think modifying existing software will be what they intend of me.
Since your modification involves adding a function which the existing unit doesn't have, the code you write will be pretty similar to what you'd get if you wrote that function from scratch, but with interesting twiddles so it works properly with the existing code to make it more difficult. It also involves understanding other people's code which is always an interesting exercise
Suggest you put it to your supervisor and see what answer you get. LeeSpeakman said:
Hi all, my name is Lee! This question may sound a bit odd, but what kind of power losses will I see by retarding the ignition and what is a safe amount?
Im involved in a university project called formula student where universities from around the world build and race single seater open wheeled racecars with a 600cc motorbike engine.
The reason I ask is im an Electronic Engineer and my project is a traction control and launch control system. One of the ways I wish to stop the wheel spin is retard/cut the spark.
The engine will probably be a supercharged 550cc aprilla V Twin, with about 80bhp @ 12000 RPM.
regards Lee S
Im involved in a university project called formula student where universities from around the world build and race single seater open wheeled racecars with a 600cc motorbike engine.
The reason I ask is im an Electronic Engineer and my project is a traction control and launch control system. One of the ways I wish to stop the wheel spin is retard/cut the spark.
The engine will probably be a supercharged 550cc aprilla V Twin, with about 80bhp @ 12000 RPM.
regards Lee S
You should aim to retard, not cut, the spark. The key is to be looking at wheel and steering sensors and then calibrate the system to allow a certain percentage of slip that yields maximum traction based on what you calibrate it to. The logic is simple - maximum traction is often gained with some slip, so you need to allow that to happen. In addition it needs to be speed dependent, because when you move off the line you do need to stop the engine bogging down. Also when you turn a corner the sensors will perceive slip, so you must account for it.
As anybody who engineers cars for a living will tell you, once you have a control strategy the hard part is calibrating it. For a road car it will take an engineer a couple of years to perfect it on one vehicle variant. But in the same way that you could spend a day getting down Santa Pod and experiment with different launch rpms to determine wheel breakaway torque in the Escort (important because you can determine throttle angle and use your dyno map to yield a reasonable result) you need to do it in the Student car so that will enable you to change all the gearing and get the system to work.
You will also need to tune the rate of cut and the rate of spark advance after the event has occurred. Again, it will need calibrating.
The hard part is going to be working out a control stategy, one that is torque based, or one that is rpm based and having something not impossible to calibrate, that also works in corners.
One of the most important things about the competition is not the sophistication of bits of the vehicle but creating something that is good overall. This means a lot of management of resources, like test time, and money to work on stuff.
I would read up on some SAE papers on controls strategies for traction control and then try to replicate it. But I would also look at what you can do to avoid the problem in the first place, which is to make the tyres work well and looking at limited slip diffs. Do not discount the idea of going diffless until you have tried it. Then there is the gearing, which needs to be matched to the torque curve and circuit.
I know that is probably not the answer you necessarily want but it is the smart route to your end goal.
Thanks for the info guys.
Im aware of the issues when turning and the engine bogging down at on a launch. I plan to have the TC cutting in at a certian threshold with all the parameters calibrated from a laptop. Im looking at measuring steering angle from the front wheel speed sensors, although i will have to look into how understeer will affect the readings.
IIRC the best amount of slip on dry tarmac is around 10%, and 5% in the wet. We will try loads of different slip levels and parameters against the stopwatch to see what works well for the car.
Im looking at using fuzzy logic for the control algorithm which is easily calibrated. before I even solder a wire the cars drivetrain and inertia will be modeled in simulink and the control algorithm tested and tuned to get it somewhere in the right direction.
Gavin, drivetrain / suspension is not my department, but im sure the guys will be doing all they can to optimize the mechanical grip / squat. At last years event none of the cars seemed to get out of 2nd gear. I hope we use a shorter gear ratio this year. As well as the TC the system will automatically shift the gears at optimum points in each gear if I get my way!
Im not sure if we are going to use a locked axle or an lsd. Last years car used an lsd so im assuming we will go down that route again. Do you think there will be much advantage going for a solid rear axle apart from weight? The event is held on a tight go-kart track at brunters.
regards Lee
Im aware of the issues when turning and the engine bogging down at on a launch. I plan to have the TC cutting in at a certian threshold with all the parameters calibrated from a laptop. Im looking at measuring steering angle from the front wheel speed sensors, although i will have to look into how understeer will affect the readings.
IIRC the best amount of slip on dry tarmac is around 10%, and 5% in the wet. We will try loads of different slip levels and parameters against the stopwatch to see what works well for the car.
Im looking at using fuzzy logic for the control algorithm which is easily calibrated. before I even solder a wire the cars drivetrain and inertia will be modeled in simulink and the control algorithm tested and tuned to get it somewhere in the right direction.
Gavin, drivetrain / suspension is not my department, but im sure the guys will be doing all they can to optimize the mechanical grip / squat. At last years event none of the cars seemed to get out of 2nd gear. I hope we use a shorter gear ratio this year. As well as the TC the system will automatically shift the gears at optimum points in each gear if I get my way!
Im not sure if we are going to use a locked axle or an lsd. Last years car used an lsd so im assuming we will go down that route again. Do you think there will be much advantage going for a solid rear axle apart from weight? The event is held on a tight go-kart track at brunters.
regards Lee
Edited by LeeSpeakman on Saturday 30th September 12:28
LeeSpeakman said:
Do you think there will be much advantage going for a solid rear axle apart from weight? The event is held on a tight go-kart track at brunters.
Talk to autotest competitors, but my guess is that the push-on understeer problems this introduces will outweigh any other benefits.
There is a bloke in the Exeter area (spotted at Wiscombe Hillclimb) with a brown Morris Minor, the thing has a 2L turbo engine and apart from being bloody fast, has a home spun traction control system. The control panel looks like a minimoog! He would be worth talking to as it was very effective.
Someone in the area must be able to put a name to him. Not sure whether it was 4WD. It was a while ago now.
Someone in the area must be able to put a name to him. Not sure whether it was 4WD. It was a while ago now.
Alright mate,
I have done formula student for the last two years, do you intend to run TC for all events or just for the acceleration event.
From my experience it is only really needed on the acceleration event, as with a decent diff (I wouldn't use a locked diff we used a torsen last year but previously have always used quaiffe ATB's) - you generally dont have enough power to break traction when cornering at a decent speed and the cars roll so little that they dont tend to pick up an inside wheel. Have seen a few cars run with solid diffs and they look like an absolute handful to drive in the sprint and endurance.
For our traction control system (I did the chassis so don't know a huge amount about this) we had hall effect sensors on one front wheel and one on the rear sproket. The ecu we used (DTA) already had a built in traction control and basically looked at the two wheel speeds and whenever there was a difference retarded the ignition to cut power until they matched again. How it actually did this in terms of software I don't know as it was a built in function of the DTA ecu.
As Gavin Pearson mentioned the hardest bit is calibrating this to work properly, we have had it on the last 3 cars but it has never worked properly due to lack of testing. In fact you are doing BLOODY well if you have the car finished with enough time to spare to calibrate it, howver if you have last years car you can use this to set it up.
To be honest the way I would do it is a lot simpler but I think works almost aswell for the acceleration event. Just set a lower rev limit that you get by holding a button down and set this to the level that when you drop the clutch you get the best possible launch with just spinning the wheels. This again needs calibrating and basically just practice dropping the clutch at set revs and find out which revs give the best start.
Good look for the event and this will be the hardest year of your educational career by a long long way!
What uni you at by the way
Alex
I have done formula student for the last two years, do you intend to run TC for all events or just for the acceleration event.
From my experience it is only really needed on the acceleration event, as with a decent diff (I wouldn't use a locked diff we used a torsen last year but previously have always used quaiffe ATB's) - you generally dont have enough power to break traction when cornering at a decent speed and the cars roll so little that they dont tend to pick up an inside wheel. Have seen a few cars run with solid diffs and they look like an absolute handful to drive in the sprint and endurance.
For our traction control system (I did the chassis so don't know a huge amount about this) we had hall effect sensors on one front wheel and one on the rear sproket. The ecu we used (DTA) already had a built in traction control and basically looked at the two wheel speeds and whenever there was a difference retarded the ignition to cut power until they matched again. How it actually did this in terms of software I don't know as it was a built in function of the DTA ecu.
As Gavin Pearson mentioned the hardest bit is calibrating this to work properly, we have had it on the last 3 cars but it has never worked properly due to lack of testing. In fact you are doing BLOODY well if you have the car finished with enough time to spare to calibrate it, howver if you have last years car you can use this to set it up.
To be honest the way I would do it is a lot simpler but I think works almost aswell for the acceleration event. Just set a lower rev limit that you get by holding a button down and set this to the level that when you drop the clutch you get the best possible launch with just spinning the wheels. This again needs calibrating and basically just practice dropping the clutch at set revs and find out which revs give the best start.
Good look for the event and this will be the hardest year of your educational career by a long long way!
What uni you at by the way
Alex
I agree with what you say and in normal cars I would tend to agree, but after driving a couple of our bike engined cars modulating the clutch is pretty difficult and you either bog down or spin too much as there is so little travel when using a foot clutch - and so little torque ......which I guess your TVR doesn't suffer from 
Edited by bales on Saturday 30th September 19:13
Edited by bales on Saturday 30th September 19:14
GreenV8S said:
What's the launch revs in terms of equivalent mph in your launch gear? If it's geared low enough you could afford to sacrifice some initial traction for an easy life.
To be honest I have no idea, I have finished uni now and it is months ago since we were working on the car and I wasn't really involved in that side of things too much anyway. Electronics(i.e traction control and stuff) always baffle me anyway I much prefer mechanical things
In all honesty when I drove the car I found that just doing it by "feel" always worked best (launching that is).
Edited by bales on Saturday 30th September 19:42
Edited by bales on Saturday 30th September 19:43
Im afraid to say that traction control is already in the MS code, I added that a few years ago, theres a basic system that looks at a rate of change in rpm or a setup that can take in 2 x VSS sensors, one from an un powered wheel and one from a powered wheel. It then look at the percentage difference is speeds and you can retard or cut sparks or even add fuel to flood the engine and reduce power, I had a little play with it on the cobra, but with other code work going on with the MS project it got put on the back burner Im afraid.
Phil
Phil
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff