Engineering FYP proposals

[tmapv]

Hi everyone,

Going into my final year this September and I'm gathering loads of ideas for a good Final Year Project (Automotive Engineering).

I want to base it on suspension or vehicle dynamics, and I could potentially do it on a Formula Student car. I want to get involved with some software such as Matlab or ADAMS with this project to build my skills with the software. I'll be more than happy to include CAD in it as well as I'm pretty good with it, but I guess this wouldn't apply if I wanted to do some Matlab based analysis or what have you... (Either way I'm open to ANY suggestion)

I'm not really sure how to utilise them within the subject. I was thinking maybe something in the lines of analysis and optimisation of a suspension system?

Any ideas would be great!

[Greg Locock]

How many hours are you going to put in? how much money do you have? can you get access to test rigs? have you any correlation data to work with? what does your professor suggest? have you got tire data?

Why not test a shock absorber made from known components on a test rig and then try and build a model of it (I believe simulink may be appropriate) that replicates some of the strange details of the test results? Shock absorber testing is cheap compared with tire testing, plus you can modify them yourself.

Plenty of material in the relevant thread here.

[krisfx]

Identify an issue with your current FSAE suspension and then work backwards from there to scope a proposal, maybe? Don't think of it as "which software can I cram in?" (Adams is horrible anyway). You should think of it as, "what problem can I solve?" My tutor also told me that good dissertations always think forward in terms of further work, so you don't necessarily have to solve the problem you first create.

Keep it simple, when it comes to March/April and you realise you've overscoped and have zero time, it becomes a bit squeaky bum, I graduated last year and those last few weeks of my dissertation (on chassis design for FSAE) were horrible!

[Greg Locock]

Yes I was going to suggest avoiding ADAMS unless you have also been through one complete cycle of build/correlate/use with it. Anyone messing about with uncorrelated ADAMS models of full vehicles is just...messing about. I'm not saying it doesn't happen, but I see no reason to encourage it.

[Greg Locock]

Incidentally optimising suspension kinematics is easy, once you know what your targets are. But your targets are set by the full vehicle, architecture (in a general sense), are entirely tire dependent, and depend on what use the vehicle has. In fact for a well defined suspension it takes less than day to get the hardpoints in the right place if the targets are compatible, which doesn't seem much of a project to me.

[tmapv]

Identify an issue with your current FSAE suspension and then work backwards from there to scope a proposal, maybe? Don't think of it as "which software can I cram in?" (Adams is horrible anyway). You should think of it as, "what problem can I solve?" My tutor also told me that good dissertations always think forward in terms of further work, so you don't necessarily have to solve the problem you first create.

Keep it simple, when it comes to March/April and you realise you've overscoped and have zero time, it becomes a bit squeaky bum, I graduated last year and those last few weeks of my dissertation (on chassis design for FSAE) were horrible!

Yeah I mean I still have to speak to the team really, I've been away for a year but had a couple bits. One of them was actually interesting. They tried out carbon fiber wishbones but they had some problems with the design and adhesives so couldn't run it. So I could take that, make another design, actually make the part and do tests on it.

But then again, my idea was to get involved in Matlab, something I want to do in the future is partially vehicle dynamics and I know that Matlab is very popular in the sector. I know that employers also look at the FYP carefully and I thought why not implement it into the FYP.

[tmapv]

How many hours are you going to put in? how much money do you have? can you get access to test rigs? have you any correlation data to work with? what does your professor suggest? have you got tire data?

Why not test a shock absorber made from known components on a test rig and then try and build a model of it (I believe simulink may be appropriate) that replicates some of the strange details of the test results? Shock absorber testing is cheap compared with tire testing, plus you can modify them yourself.

Plenty of material in the relevant thread here.

It's worth 2 modules so probably 20 hours a week, don't know really just a guess. I have saved up a bit, shouldn't be a problem as long as I don't have to build a car or something And yeah we have some test rigs, although would have to confirm what exactly as I can't remember due to being away for a year but the basics like tensile testing, car sensors etc. I could relate a lot of it to our previous Formula Student car, and yeah we also have tyre data. How could I use the tyre data in matlab to develop a suspension system? One of the ideas was to do analysis of passive vs active suspension, could I maybe implement the tyre models into that?

And the shock absorber one sounds good too, I think some hands on stuff would be good but I wanted to include matlab because I need to develop my skills in it. I'm sure skills in the software is something employers look at especially for suspension design/vehicle dyanmics.

[krisfx]

Identify an issue with your current FSAE suspension and then work backwards from there to scope a proposal, maybe? Don't think of it as "which software can I cram in?" (Adams is horrible anyway). You should think of it as, "what problem can I solve?" My tutor also told me that good dissertations always think forward in terms of further work, so you don't necessarily have to solve the problem you first create.

Keep it simple, when it comes to March/April and you realise you've overscoped and have zero time, it becomes a bit squeaky bum, I graduated last year and those last few weeks of my dissertation (on chassis design for FSAE) were horrible!

Yeah I mean I still have to speak to the team really, I've been away for a year but had a couple bits. One of them was actually interesting. They tried out carbon fiber wishbones but they had some problems with the design and adhesives so couldn't run it. So I could take that, make another design, actually make the part and do tests on it.

But then again, my idea was to get involved in Matlab, something I want to do in the future is partially vehicle dynamics and I know that Matlab is very popular in the sector. I know that employers also look at the FYP carefully and I thought why not implement it into the FYP.

My FYP was looked at once in all the interviews I went to, and it wasn't the automotive one I took, haha.

[tmapv]

I know but as it''s my own personal project and study, I want to use it to develop my abilities in the particular subject. Even if they don't mention it, why shouldn't I?

[krisfx]

I know but as it''s my own personal project and study, I want to use it to develop my abilities in the particular subject. Even if they don't mention it, why shouldn't I?

As was mine, but what I'm trying to get across is that it's almost as much about demonstrating your project management ability as it is about content

[Tim.Wright]

The BIIIIG thing missing in most FSAE projects regarding vehicle dynamics is any quantification of the car's handling performance. I mean more than just saying "the car reaches 7g on the skidpad" which is useless information on its own.

So to resolve that problem you should develop, for your team, a methodology of simulation and post processing which quantifies the grip performance, stability, control-ability etc. The best way to do this would be to implement a program in Matlab which post processes simulation and test data from standard open loop steering tests (forget lap data for now).

I.e. What is when some judge tells you that your car is wrong because you have your LLTD so different from your CG location, you can show that your static margin, lat acc response time and and lat acc delay suggest otherwise. When another judge tells you that he doesn't like your rear toe mount because it looks too flexible and it will make the car's response sluggish - you can show the response to a step steer test together with all the dynamic metrics to prove that the car is doing what you want. When someone asks how you chose your tyres you can show dozens of metrics calculated from simulations using the TTC data which ranks the tyres and more areas than you would have time to explain.

In short - some sort of analysis environment like this is the difference between engineering a car and just hacking something together.

A good source for all this stuff is Bill Cobb's posts on fsae.com. Bill is an ex GM vehicle engineer who is quite active in spreading the word (via matlab code) on how to objectively quantify vehicle handling as opposed to the handwavy stuff you read in most motorsport books or seminars. Books by Milliken, Guiggiani, Pacejka & Mitschke will also help. Avoid most of the others - particularly motorsport specific ones because they are the worst from the "in-quantifiable handwaving BS" point of view.

[tmapv]

The BIIIIG thing missing in most FSAE projects regarding vehicle dynamics is any quantification of the car's handling performance. I mean more than just saying "the car reaches 7g on the skidpad" which is useless information on its own.

So to resolve that problem you should develop, for your team, a methodology of simulation and post processing which quantifies the grip performance, stability, control-ability etc. The best way to do this would be to implement a program in Matlab which post processes simulation and test data from standard open loop steering tests (forget lap data for now).

I.e. What is when some judge tells you that your car is wrong because you have your LLTD so different from your CG location, you can show that your static margin, lat acc response time and and lat acc delay suggest otherwise. When another judge tells you that he doesn't like your rear toe mount because it looks too flexible and it will make the car's response sluggish - you can show the response to a step steer test together with all the dynamic metrics to prove that the car is doing what you want. When someone asks how you chose your tyres you can show dozens of metrics calculated from simulations using the TTC data which ranks the tyres and more areas than you would have time to explain.

In short - some sort of analysis environment like this is the difference between engineering a car and just hacking something together.

A good source for all this stuff is Bill Cobb's posts on fsae.com. Bill is an ex GM vehicle engineer who is quite active in spreading the word (via matlab code) on how to objectively quantify vehicle handling as opposed to the handwavy stuff you read in most motorsport books or seminars. Books by Milliken, Guiggiani, Pacejka & Mitschke will also help. Avoid most of the others - particularly motorsport specific ones because they are the worst from the "in-quantifiable handwaving BS" point of view.

Okay sounds interesting . How would the process look? I guess first of all, I would need to get some tyre data out of the one we chose. For calculating the dynamics say round a corner I would need then all of the vehicle parameters, so a lot of the design has to be established already doesn't it?

Also, is all of this possible in Matlab? I presume I would need to run a full vehicle model for this and I thought that Matlab was not the most efficient at this, let alone possible at all? I havn't seen anything like this done in Matlab anyway, but I'll keep looking...

[Tim.Wright]

What I was proposing was not to write a vehicle simulation in Matlab (though this is totally possible and has been done many times before). I proposed that you write a program that takes simulation or test data and analyses it in order to quantify the vehicle's handling performance.

All of the major vehicle manufactures have software like this developed in house and some students have done this already in FSAE.

Have a look at the ISO/SAE standard for:
Steady state cornering
Transient response
Power on (actually not a huge fan of this method but useful as a starting point)
Power off (not a huge fan)

Your program would then take the test data, perform the necessary sensor processing (filtering, corrections, transformations etc) and then calclulate various metrics as defined in the ISO norms, Milliken, Guiggiani etc..

That way, you can quantify all your decisions - both to appease design judges as well as fellow team members.

Otherwise - what else are you going to say when someone asks you why have you chosen this much bumpsteer? Why have you complicated your life with push rods and rockers? Then there is usually a team member telling you that beam axles are the way to go (they are probably right too). How else can you answer these questions or make these decisions in an informed way if you have no way of quantifying the handling of the car?

Like I said before its THE biggest thing missing in FSAE at the moment. Chassis/suspension choices are almost always made with an equal mix of rules of thumb and handwaving which is the polar opposite of engineering.

[Tim.Wright]

Alternatively, if you want to learn adams - your project could be the construction AND VALIDATION of a handling model.

[krisfx]

Completely agree with Tim.

I noticed in some FSAE teams that design decisions were also made "because that's how they do it in F1(or other generic series)" and as Tim said, a lot of it ends up just being who can shout the loudest. The first part of my FYP was to devise the methodology to rule out the handwavers and to quantify necessary data to ensure that the choices made were more logical.

I always wondered what it would be like to run as stiff as a kart and whether that would be as quick/quicker