Active Rear Wing Control System. Help Needed

First off a quick introduction,
Although I'm not a builder of robots myself I am very interested in the operation and controls of electronic actuators. I'm a mechanical engineer (in training) living in Dublin in my final year of my course and I'm midway through completing my thesis project which I hope some of you here can help me with.

In order for those interested to understand the idea behind this project the following text is required so apologies for the long read. I'm sure some of you will get bored half way through and I don't blame you as its very wordy! :)

So the title of my project is "The Design and Construction of an Active Rear Wing for an Alfa Romeo Touring Car".
The active wing system system is for a 3.2l V6 Alfa Romeo 155 touring car that will eventually be used in the Irish Touring Car Championship. As said before the project is for my final year in college and it will all be designed and built by myself. The wing will be moved by two GL-750s Linear Actuators (with 50mm of travel) which are mounted in the boot vertically. These are connected to the wing by link rods which change the wings angle of attack. The wing is required to have three preset positions A, B and C.

Postion A is the Low Drag DRS position at which the wing is at 1 degrees of attack. This will give optimum straight line speed.

Position B is the High Downforce position at which the wing is located at 15-18 degrees of attack. This will give optimum levels of downforce while cornering.

Position C is the High Drag Air Brake position which creates maximum drag in order to slow the car while under braking.

So the idea is that the wing will move to these various positions semi automatically, I have an ideal system that sounds fairly straight forward in theory but I am really lost as to how the electrics will work so I am open to any ideas or suggestions.
As said earlier the system uses two actuators which must work in sync. The actuators been used have no feedback or limit switches as standard so these need to be incorporated externally. They will run off the cars 12v battery which is also located in the boot.

Position B will be the wings home position as such.
The steering wheel has a single push button which the driver pushes when he comes onto a straight section of track. This then activates the actuators to move to position A.
When the driver comes to the end of the straight he then brakes as usual. The brake lights will then override the button control and activate the wing to move to position C which slows the car before the corner. Once the driver takes his foot off the pedal the wing then moves back to its home position.


I know its not a simple theory so at the moment i would be happy just to get some sort of control on the wing and then move from there. I really would appreciate any feedback or advice. Some lectures have mentioned the use of Arduino hardware but I have no experience with this.

The plan is to have the wing built in the next few weeks and then from there i will move onto the electonic controls

Thanks for taking an interest and making it to the end of this!

Below is a basic design although no link rods are shown here.
http://www.cloverleaf4.co.uk/images/uploads/29465.jpg

Thanks again
Peter

I agree with the Arduino. Its going to give you flexibility in the control mechanisms and algorithms. They are simple to program and there is alot of advice out there - http://forum.arduino.cc/
Good luck on the project

If you use an Arduino or something similar, you might be able to use accelerometers to automate the wing; one sensor could detect sideways motion to move between positions A & B. a second sensor could detect deceleration to move to position C. The nice thing about that is you could get proportional control with little driver input needed.

Another thought; with two actuators it would be quite easy for them to get further & further out of sync if one of them ran just a bit slower than the other, leading to damage or lock-up. If you use an Arduino, you can use one limit switch to set a home position and a rotary encoder on each actuator to make sure they are in sync and get full proportional control, which seems like a good idea for braking.

I couldn't find the actuators on Google, do you have a link for them?

As a pure mechanical guy I can construct a relay-switch based setup, but I fear it will fill the trunk.

No, the arduino sounds like the best bet. And Nicks comment are spot on.

To get perfect sync, use 2 encoders , or even good quality potmeters that measure the axle position of the wing-pivots.

If the idea of measuring acceleration in 2 dimensions is workable, you can even have the wing, if it's flexible enough, torque to have more or less downforce at each side. You could even go further, and do the same with the front spoiler.

I was thinking about an encoder on the wing itself but the what if the actuator on the far side ran at a different rate? It would still twist the wing and you'd never know. With just one actuator it would be fine.

I wonder if an actuator will move fast enough to position the wing in race conditions? The heavy duty actuators I have seen are pretty slow and race conditions change very quickly. They would be the simplest off-the-shelf parts to use but if they don't deliver, they aren't worth anything.

Another place to look for parts might be in the CNC world. A servo motor and an ACME threaded rod or ball screw would move much faster and servo motors usually have an encoder built in. Looking back at Peter's original description, the system already uses pushrods up to the wing, so I would look at a nut up on the wing with a short length of screw and then a rotation shaft down the the trunk. More effort to make, but fewer moving parts, probably lighter and it could move the wing much faster.

Another limitation is the 12V supply; no over-volting anything to get more speed or torque.

Encoders all the way only way to be sure on the sync of the motors,all other option wouldn't have fine enought control to have the wing perfectly square . Also you can then have the wing linked to the Speedo so it adjust continuesly with the vehicles speed

The 12 volt supply ain't an issue. Replace the extra lead battery with a LiPo pack of the same weigth, and a 6s psck is cheap and rechargable with a equaly cheap and light charger that happely will use the on board 12volt supply.

But yes, you're right Nick, a normal linak is artritic slow for this purpose.

With only 3 positions I see a solution with 4 solenoids.
How fast is an Arduino? The speed of a mosfet is known and more than fast enough.
This also solves the sync issue.

An Arduino is defiantly fast enough. You can make it output its readings at up to 115'200 per second. Making it do that process actually slows it down so if you take it out it will cycle through the code as fast as it can.

I would say lock it to either 2400, 4800 or 9600 Baud depending on how accurate you want to be. It is far harder to diagnose issue as high baud rates so make life easy on yourself and go for something manageable then crank it up in the future if you need to.

Some Arduino boards are faster than others; I have never used them but there are ARM 32 bit Arduinos with serious processing power like this: https://www.sparkfun.com/products/11589

I don't know if solenoids would have the power to shift the wing while the car is in motion - I recall hearing that F1 car wings generated enough down-force to allow the cars to drive upside down, so a touring car wing must generate hundreds of Kg of down-force.

hydraulics?

Wow, genuinely amazed at the response to this. Thanks to everyone for the replies and input. Plenty of interesting ideas floating around there.

To put things in context, this project is already far beyond what is needed for this thesis and although it sounds counterintuitive I need to try and keep the electronic controls as simple as possible. Reason been the course I am completing is mechanical engineering so too much focus cannot be placed on the electronics alone. In regards to having progressive control there's no real need for it. As under braking the wing needs to move as quickly as possible. And then back to high downforce once the braking is finished. Also, hydraulics cannot be used as its a safety issue. The forces involved at the high downforce levels will be about 1500 newtons. As for when the wing moves to air brake mode, as the wing is in stall this needs to be calculated via a wind tunnel test although I would expect more that 2000N.

http://www.gimsonrobotics.co.uk/GLA-..._actuator.html
Here's the actuators been used, as said in the description they can move at up to 30mm/second. Although thats is far from quick the problem is overcome by having the link to the rod right next to the pivot point. So with a very short extension the wing can change angle quite severely.

One simple idea i had was to use two limit switches which would be activated by a pin mounted on the lower part of the connecting rod beneath the boot lid.

So when the button on the wheel is pushed the wing is activated to move in the positive direction as such and moves upwards until it reaches the upper limit switch at which point it stops.
Then when the brake pedal is pushed, maybe just through the brake lights a current could be taken to activate the actuators to move in the negative direction, once they reach the required air brake angle and trip the limit switch then the actuators stop again.
However this still leaves the problem of how to set a home position for when neither an input from the button or brake lights are received.

Quote:

---

Originally Posted by overkill

If you use an Arduino or something similar, you might be able to use accelerometers to automate the wing; one sensor could detect sideways motion to move between positions A & B. a second sensor could detect deceleration to move to position C. The nice thing about that is you could get proportional control with little driver input needed.

---

As mentioned above there's no real need for proportional control with this system as it basically can be said there is just positionA, B or C. Pick one! haha. Although there could be a system with proportional movement it would be far more complicated than what I'm at.

Quote:

---

Originally Posted by overkill

Another thought; with two actuators it would be quite easy for them to get further & further out of sync if one of them ran just a bit slower than the other, leading to damage or lock-up. If you use an Arduino, you can use one limit switch to set a home position and a rotary encoder on each actuator to make sure they are in sync and get full proportional control, which seems like a good idea for braking.
I couldn't find the actuators on Google, do you have a link for them?

---

Only this morning I was questioning my choice to use two actuators, it is a problem. But the two are needed as the forces using the airbrake could be upwards of 2000N so there needs to be an element of factors of safety. Surely if there are limit switches though set in the same physical position then they would stay in check?

Quote:

---

Originally Posted by maddox10

As a pure mechanical guy I can construct a relay-switch based setup, but I fear it will fill the trunk.
No, the arduino sounds like the best bet. And Nicks comment are spot on.
To get perfect sync, use 2 encoders , or even good quality potmeters that measure the axle position of the wing-pivots.
If the idea of measuring acceleration in 2 dimensions is workable, you can even have the wing, if it's flexible enough, torque to have more or less downforce at each side. You could even go further, and do the same with the front spoiler.

---

What do you mean by it would'nt fit the trunk?
To be honest the wing is very much over engineered for safety reasons so there will be little or no twist.
Once I am done this years college I'll be moving onto the rest of the aero bits on the car, it really is endless the amount you can do but its incredibly interesting. Bits will include a front splitter, new cooling system, flat floor and a diffuser. So plenty of work to be done, in the future

Quote:

---

Originally Posted by overkill

I was thinking about an encoder on the wing itself but the what if the actuator on the far side ran at a different rate? It would still twist the wing and you'd never know. With just one actuator it would be fine.
I wonder if an actuator will move fast enough to position the wing in race conditions? The heavy duty actuators I have seen are pretty slow and race conditions change very quickly. They would be the simplest off-the-shelf parts to use but if they don't deliver, they aren't worth anything.
Another place to look for parts might be in the CNC world. A servo motor and an ACME threaded rod or ball screw would move much faster and servo motors usually have an encoder built in. Looking back at Peter's original description, the system already uses pushrods up to the wing, so I would look at a nut up on the wing with a short length of screw and then a rotation shaft down the the trunk. More effort to make, but fewer moving parts, probably lighter and it could move the wing much faster.
Another limitation is the 12V supply; no over-volting anything to get more speed or torque.

---

As said above they actuators used can move at 30mm/sec under load. To optimise the speed the connecting point on the wing will be located very close to the pivot point.
Although its not ideal it will do the job.

Quote:

---

Originally Posted by adz600

Encoders all the way only way to be sure on the sync of the motors,all other option wouldn't have fine enought control to have the wing perfectly square . Also you can then have the wing linked to the Speedo so it adjust continuesly with the vehicles speed

---

Severaly people have mentioned the gradual control before it raises some other issues as I like the idea of having an always home position as it acts as a safety as the wing always returns to the high downforce mode, giving optimum control unless another input is recieved form the push button or brake lights

Quote:

---

Originally Posted by Niels - Team K.a.L.M.

hydraulics?

---

As said above, hydraulics were ruled out as they raise a safety issue

Also for those of you with an interest in cars, here's the car that the wing is been fitted to...

https://www.youtube.com/watch?v=bSRw0nDB7qc

mr giuseppe busso's best effort :) My dad is a race mechanic so I appreciate that engine

Hi Peter, I'm sorry if we are over-complicating things for you - its in our nature to come up with unique solutions and improve on whatever we are working on.

Although I can see the reason for keeping the electronics as simple as possible, using an Arduino or other micro controller (MCU) will likely be simpler than just trying to do it with relays. By having each input (buttons & switches) and each output (the actuators) monitored & controlled separately by the Arduino, you get more flexibility to fix problems and get the system working the way you intended. Even without all those accelerometers and encoders, an MCU is still the best solution.

Looking over your posts, you have eight inputs: three switches per actuator, brakes and a button. There could be one or two outputs and I would advise two so you can control each actuator separately to overcome any speed/position differences. As well as these basics, I'd recommend one or two extra outputs to give the driver some feedback - he would definitely want to know if the wing was stuck in brake mode! The system only has three states. That is a really easy situation for an MCU and would be handled by a small logic table, which we can work out for you if you are not sure about it. You mentioned safety was a big issue and an MCU can help by monitoring and reporting faults (maybe just a failure light on the dashboard).

Those actuators might not be what you expected. Looking at the specs, the 31mm/sec speed is with no load. From experience and reading other spec. sheets, the actual speed at about half it's rated load with be much slower. Have a search for higher-end actuators and you will often find graphs showing their speed VS load - its pretty disappointing sometimes! If you look at the GLA4000 actuator, it mentions a 40% speed reduction at full load and it probably has a much higher gear reduction than the GLA750s. I'd expect the GLA750s to slow down by a larger percentage under load.

Actuators usually state a duty cycle rating and its often quite low. If the wing is moving frequently, you might find the actuators overheat.

The 550 motors may not be reliable under heavy loads and they are definitely not weatherproof, so the trunk will need to be. They could also be a fire hazard as the brushes produce a lot of sparks and the open cans will allow fuel vapour in. The type of sealed motor on the GLA4000 is what you probably need. I realise the GLA4000 doesn't meet your other requirements, but that construction style and IP65 environment rating is the way to go.

After looking at the drawings and your comments, I have a couple of questions:

What part of the car are the wing supports mounted to; the trunk lid or the body?

You mentioned up to 2000N force. Is that the total on the wing or on the point where the actuators attach? This is a class 3 lever arrangement and if the 2000N is considered to be applied on the trailing edge of the wing, you will need considerably more force to elevate the wing into brake mode. I know nothing about aerodynamics, thought I'd better ask.

Are you up for some mechanical work? Would modifying an actuator to get what you want be possible?

What about the electronics work? Could you solder a few components together of do you need everything to be plug & play?

Hi there,

I ment that an old school end switch/relay based setup would be large and heavy, offsetting most advantages the idea can have. One of the reasons active wing control ain't that much used.

After looking several youtube's with cars with active wing control I'm almost sure the GLA linaks will come up short.

These just don't have the grunt/speed for the demands imposed on them for an effective active wing.

Nicks remarks stay to the point. I'm curious how custom build you want to go. Building 4000Nm linaks with speeds up to 1m/sec isn't rocket science.

Have to agree about a custom actuator and it might not be much work. Using that GLA4000 as an example, it only has a 50 watt motor and I'd bet you a pint of Guinness that is fairly low RPM as well. By changing the motor for one with 4 or 5 times the RPM and maybe 100 to 200 Watts power, you would get the required force and speed without having to make an actuator from scratch.

Some random thoughts:

Using two actuators is probably LESS reliable than using one larger one. If a motor or gear fails, it will lock the actuator solid and the other actuator will be trying to twist the wing - either something mechanical will break of the 2nd motor will burn out quickly.

If you use an Arduino, this motor control shield might be a good match:http://tinyurl.com/bmp4lmy. It uses automotive motor control chips that have temperature & current limiting and should handle high power actuators. It will work with very little programming and has advanced features you could use at a later date.

What sort of budget are we working with - penniless student or rich sponsored/factory team? :)

The forces acting on the push rod and actuator are always going to be compressive, which is exactly how actuators are meant to be used. For the push rod, I'd be using a wider tube rather than a thin solid rod as it will deflect less in compression for the same weight. Add a million style points if you can use a carbon fibre tube.

Quote:

---

Originally Posted by overkill

Have to agree about a custom actuator and it might not be much work. Using that GLA4000 as an example, it only has a 50 watt motor and I'd bet you a pint of Guinness that is fairly low RPM as well. By changing the motor for one with 4 or 5 times the RPM and maybe 100 to 200 Watts power, you would get the required force and speed without having to make an actuator from scratch.

Some random thoughts:

Using two actuators is probably LESS reliable than using one larger one. If a motor or gear fails, it will lock the actuator solid and the other actuator will be trying to twist the wing - either something mechanical will break of the 2nd motor will burn out quickly.

If you use an Arduino, this motor control shield might be a good match:http://tinyurl.com/bmp4lmy. It uses automotive motor control chips that have temperature & current limiting and should handle high power actuators. It will work with very little programming and has advanced features you could use at a later date.

What sort of budget are we working with - penniless student or rich sponsored/factory team? :)

The forces acting on the push rod and actuator are always going to be compressive, which is exactly how actuators are meant to be used. For the push rod, I'd be using a wider tube rather than a thin solid rod as it will deflect less in compression for the same weight. Add a million style points if you can use a carbon fibre tube.

---

Hi all,
Apologies for the lack of replies, I have been mad busy lately.
I received the actuators, they are very very small and light.
After considering a lot of your input (and others from various backgrounds) i have decided to make several changes to the design.

1. Use only one actuator, it seems this is a logical step for reliability and also aerodynamics. The actuator I have can handle 2300N so that will be ok on its own.
In the future i can upgrade my motor to be more powerful/quicker

2. The actuator and wing will be mounted to a subframe as such in the boot. The boot lid will play no role in suporting the wing, it will just act as a cover. The car has gone to get a role cage this week so the work will be completed then.

3. The actuator will act in the centre of the wing between the two mounts.

4. The wing has been increased in size to 1700x400mm
Here are the latest calculations after these changes...
http://www.backroads.ie/forums/attac...844188&thumb=1
Thats 125kg @100mph! Happy days


I'll try to keep here updated, thanks all for the input, it has been an eye opener to some aspects. I hope to be working on the mechanical construction for the next 3 weeks so I won't be focusing on the aspect at hand here. But once the wing is completed I'll have photos up for all to see and then I can concentrate on the controls aspect.

Thanks again,
Peter

Also, @overkill
I am most definitely a poor student haha

Hi all,
I have the wing nearly built now so its nearly time to move onto the control side as my main focus.
I have re read all of the posts here just to make sure I am up to date.
First off, I have decided to use only one of the actuators. So that will reduce some of the work load.

Assuming I am to use an Arduino or othe MCU what do I need otherwise.
The actuator I have chosen (maybe foolishly) have no feedback or limit to stop it from running beyond its maximum range. So a potentiometer or some sort of feedback is required? Is this then feed into the MCU? Just trying to get an idea of what more I will need to complete this project

As you can probably tell I am not much of an electronics head so apologies if I seem totally lost here.

It would be better to use a pair of microswitches rather than a potentiometer - they are easier to interface and much less affected by heat and dirt. You will need to find a convenient spot to mount the switches, probably on the actuator or nearby. You arrange the switches so that the actuator switches them on at each end of it's travel.

What might be even easier to set up is three switches that indicate the positions of the wing since the up & down positions of the wing are essentially the same as the actuator's limits. The middle switch indicates the neutral position and you just need to adjust the switch positions on the bracket to fine-tune the wing position.

The switches will be connected to three digital input pins on the MCU and the software will use the switch inputs to know when the actuator is in the right position.

Quote:

---

Originally Posted by overkill

It would be better to use a pair of microswitches rather than a potentiometer - they are easier to interface and much less affected by heat and dirt. You will need to find a convenient spot to mount the switches, probably on the actuator or nearby. You arrange the switches so that the actuator switches them on at each end of it's travel.

What might be even easier to set up is three switches that indicate the positions of the wing since the up & down positions of the wing are essentially the same as the actuator's limits. The middle switch indicates the neutral position and you just need to adjust the switch positions on the bracket to fine-tune the wing position.

The switches will be connected to three digital input pins on the MCU and the software will use the switch inputs to know when the actuator is in the right position.

---

Thanks for the reply overkill,
I had initially thought i could go about it that way as it seemed straight forward and simple. It might be tricky to make up a mounting bracket to house the microswitches. I could have a small tab as such on the connecting rod from the actuators which as you say, trips the limit switches one by one. One difficulty is that the movement of the actuator is only 50mm so the switches would need to be mounted quite accurately.
Would anything else be required other than the limit switched and an MCU which can take 5 inputs?

Not much; the main items will be a couple of relays (or a single DPDT relay) and a simple circuit to drive them as an MCU output doesn't have the capacity to drive a relay directly. Alternatively, you can use an 'H bridge' which essentially does the same job as the relays but its all electronic with no moving parts. You will also need a DC to DC power supply to drop the car's 12 volts down to 5V or 3.3V for the MCU. almost all that stuff is "off the shelf", I or the other furum members can probably whip up a circuit and a shopping list if you need it.

Quote:

---

Originally Posted by overkill

Not much; the main items will be a couple of relays (or a single DPDT relay) and a simple circuit to drive them as an MCU output doesn't have the capacity to drive a relay directly. Alternatively, you can use an 'H bridge' which essentially does the same job as the relays but its all electronic with no moving parts. You will also need a DC to DC power supply to drop the car's 12 volts down to 5V or 3.3V for the MCU. almost all that stuff is "off the shelf", I or the other furum members can probably whip up a circuit and a shopping list if you need it.

---

That sounds great, if I could see what I actually need it would be easier to work from and also easier to approach lectures in college if I need help as then I'd have something to work with. Really appreciate the help

No Problems, I need a bit of a break from robots today. Do you prefer to control the motor with relays or with electronics. The main differences are that the relays will be a bit cheaper but will need more custom wiring. Using a solidstate H bridge might be a bit more compact as it can fit right on the arduino board

Quote:

---

Originally Posted by overkill

No Problems, I need a bit of a break from robots today. Do you prefer to control the motor with relays or with electronics. The main differences are that the relays will be a bit cheaper but will need more custom wiring. Using a solidstate H bridge might be a bit more compact as it can fit right on the arduino board

---

This is a break? haha

Correct me if I'm wrong anywhere here!

Thats 2 options then:
1. Relays using physical contact as feedback as such? Seems fairly simple to set the limits of movements/set points
2. H bridge, how are the limits of movement set? and the set points, this is where my total lack of knowledge towards electronics is blatantly obvious!

But in both cases i need:
1. DC to DC power supply
2. Arduino

Any ideas of a very vague cost to get this system up and going? Projects cost about €300 so far

The motor drive with relays or an H-bridge is completely separate from the position sensing, with the MCU in the middle to take the position and driver switches and translate that into turning the motor in the right direction. All the relays (or H-bridge) does is power the motor; they don't have anything to do with feedback.

The DC supply would just be 3 or 4 small components to step the 12V down to 5V; they can be soldered onto a small piece of veroboard with basic skills or a pre-built item can be sourced (we call them a 'BEC' for battery eliminator circuit). Since the project has few inputs and outputs, you really only need the most basic Arduino board probably the Arduino Uno at around 18 pounds. I don't know many suppliers in the UK, so unless you have a favourite, I will use Element 14 or RS-Components as a reference.

Before I go designing the wrong solution, are you up for any electronics soldering? The relay option is looking cheaper than than the H-bridge but its more work to construct, while the bridge is very plug & play.

Quote:

---

Originally Posted by overkill

Before I go designing the wrong solution, are you up for any electronics soldering? The relay option is looking cheaper than than the H-bridge but its more work to construct, while the bridge is very plug & play.

---

Hi again overkill,
I would be able to solder no problem but i dont have much time so plug and play is the way to go!

chatting to a lecturer today in college and he reckon arduino is the way to go, as it will be the most straight forward option for me so i dont get too tied up in the electronics. also recommended using a motor shield, im heading out to a supplier today to hear what they have to say.

What do you think of the arduino and motor shield option?

Got the arduino uno and motor shield.
This motor rus at 32A at max load and 72A so i think i will need to operate a relay with the arduino?
any ideas?

Sorry I haven't replied much, I just flew in for the UK robot championships. I was thinking of using a motor shield that would drive the actuator directly, but a smaller one should drive relays just fine. I realized that it will need three relays - 1 to switch the power and two to provide the forward & reverse. Can you post a link to the motor shield you bought?

Quote:

---

Originally Posted by overkill

Sorry I haven't replied much, I just flew in for the UK robot championships. I was thinking of using a motor shield that would drive the actuator directly, but a smaller one should drive relays just fine. I realized that it will need three relays - 1 to switch the power and two to provide the forward & reverse. Can you post a link to the motor shield you bought?

---

Right back at you with the lack of replies!
here ya are.
Soooooo my idea now.

Battery-Arduino-motorshield-50A relay-two limit switches
So basically reducing it to two positions, low drag and high downforce. Push button to move wing up into low downforce position, meaning it moves till hits limit switch.
Then second position is activated by brake pedal moving actuator in reverse until it hits the lower limit switch.
Just seems the simplest solution overall.

Here ya are.
Motor shield
http://radionics.rs-online.com/web/p...382D3933343926

Arduino
http://radionics.rs-online.com/web/p...392D3734303926

My thesis is due in tomorrow. Here's where I'm at..
https://www.youtube.com/watch?v=8G_VMzmjxkc

https://scontent-b.xx.fbcdn.net/hpho...09292251_n.jpg

https://fbcdn-sphotos-a-a.akamaihd.n...25833269_n.jpg