My questions about camshaft design

[PedroCG]

Hi, the past month or so i've been diving in this subject, i've read some Race Engine Technology articles by prof Blair and Associates and whats available trough google books of Norton's "Camshaft Design and Manufacturing".
My objective is to develop an accurate enough software to predict valve train dynamics for a direct acting (bucket) system. The software will be used to project a cam for a FSAE car.
I'll use this thread to get some help in some questions.

The first thing i looked at was those 3 arc cams then polynomials and then splines. Splines prove to be the best, all polynomials will work but some will be more troublesome than others to "fine tune".

So my first question is: whats the problem with infinite jerk? 3 arc cams have infine jerk at the inflection points, they also have less area under the displacement for higher acceleration values and those are clear disavantages, but the jerk part i dont understand since its not related directly to force it will not cause loss of contact. i guess

 

[ivymike]

why don't you just get a student license of commercial software like ValDyn?

 

[PedroCG]

1º - I was suppose to be taking a BSc in automotive engineering turns out the school hasn't tougth me nothing automotive yet (second year) so i better learn things by myself.

2º - Student license doesn't mean free, at least thats the case for 4stHEAD and solidworks.

3º - I have this theory about learning, when someone teaches you something you... learn, when you understand something by yourself you get smarter.

sorry if this "sounded" bitter, but i'm realy disapointed with school.

 

[SomptingGuy]

There are often subtle differences between pure academic (i.e. student) licenses and those granted for FSAE use.

- Steve

 

[facty]

How long have you got!? Taking on this type of project does not normally fall under the remit of a 'student project' - I don't mean you can't do it, but you'll probably find that you won't have the time or resources to do it right.

The commercial softwares you've mentioned (Valdyne, 4stHEAD etc) that tackle valvetrain dynamic analysis cost significant money because many hours and hundreds of thousands of lines of code have been used to perfect them - not to mention time spent on validating the algorithms used with special test rigs or actual engine testing!

Look at the complexity of the mathematical model required to attempt a dynamic analysis on a bucket valvetrain:

http://www.profblairandassociates.com/GPB_4stHEAD_ValveTrainAnalysis_theory.html

The analysis must include bounce and separation between all the parts and also clash between the coils in the spring - this is not easy to accomplish mathematically. What you will also find is that in attempting to do this there is very little published information on the damping coefficients used - and it is these that can greatly influence the ensuing solution.

The main input to such an analysis is the valve lift profile - again, as you have mentioned, this is no easy topic to tackle. The commercial softwares use combinations of curves, splines and smoothing techniques that take many thousands of lines of code to perfect - a project in itself!

To answer your question re jerk - think of jerk like an impact - if your jerk is high then you may as well be hitting your valvetrain with a sledge hammer.

Finally, you might find that if you contact the Sales guys supplying these commercial softwares that they have "freebie" policies for FSAE teams!

 

[Tmoose]

http://books.google.com/books?id=2RQkAAAAMAAJ&source=gbs_ViewAPI&pgis=1

I believe Acceleration and jerk translate into extra load on the valve train. If the "contact stress" on the cam follower exceeds some nominal value (depends on materials) then wear and fatigue become problems.

Non smooth acceleration or pulses can provoke resonant frequencies. If the resonance is spring surge then spring reliability is directly compromised and valve bounce can appear. That's no fun.

 

[PedroCG]

facty, i understand the complexety of the analysis. my next step is to study MDOF systems, after that i'll be able to calculate oil film tickness and bucket stress. about valve lift, im using two B-splines with 10 points each (one for opening other for closing) first and last 3 points on each curve set lash, ramp velocity, max lift and negative aceleration at the center, the other 4 points are moved to manipulate the displacement curve, i can replicate the acceleration and velocity profiles of the designs in prof Blair's articles with same KLD values. I migth need to add one more point to control max jerk as that is a bit high. i can also translate the valve lift profile to the physical form of the cam lobe.
damping coeficients will be dificult to find but... one thing at its time.
for now im satysfied with the jerk explanation, its hard to look at it separately from acceleration, i guess water inside a bottle is a good example, with high jerk the water goes crazy inside the bottle.
Tmoose, resonace will allways be a problem, theres so many harmonics that its hard to avoid them, one just shoots for a range of frequencys where the harmonics are low.

thanks for the replys.

 

[facty]

The coding of the MDOF is quite a challenge if you hope to model bounce, separation and clash within your simulation. I'd suggest using a Runge-Kutta approach to solving your MDOF system. Start with a simple system - no bounce, separation or clash - get this working with a robust solver and then deal with the more complicated scenario.

 

[facty]

Pedro, here's an FSAE paper that describes an FSAE project to redesign the camshaft to improve performance - might be of some interest:

http://www.sae.org/technical/papers/2008-32-0073

 

[PedroCG]

hi, thanks again facty, i've founded that paper a while ago, i haven't bougth it yet because we are going to turbocharge the engine.

 

[larrylcoyle]

PedroCG,

Have you considered partnering with one of the camshaft developers at one of the more reputable camshaft companies? If you are interested in going that direction, I would be willing to make the necessary introductions and I can assure you this guy is as smart as anyone in the country that is developing camshafts. One thing though, he does not use commercial code and if I told you the code he uses for his development, you would probably fall out of your chair.

Another thing, what are your rules/limitations as far as cylinder heads? Is CNC Porting allowed? If so, and you have someone on your team sharp enough to develop the port designs, I would be willing to donate my time and my machine time as well since that is my business.

Larry

 

[patprimmer]

Larry

Have you read the site rules and guidelines. You have promised to break the rules in your 2 recent posts and you are VERY borderline in this post. To avoid red flags, I suggest you read, comprehend and abide by the rules.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[PedroCG]

larry, i think theres no rules about head porting but there's rules about outside services, it must be a student project... i would consider partnering after i see the results of what im doing, i live in Portugal so i dot know how many people are developing (in a serious way) camshafts.

 

[larrylcoyle]

Pat,

While it was not my intention to do so, I see what you are talking about and I wish to thank you for bringing this to my attention and keeping me honest. This will not happen again as I will be more careful regarding any post of mine.. I truly did not think that volunteering a service was a violation though for someone such as a student.

Larry

 

[patprimmer]

Larry

That is why I posted advice rather than just red flagging it.

In my fairly long experience here, judging by posts that stay vs those that disappear, if you do post e-mail addys it will get red flagged. You can post links to a URL which then contains contact details if you do so in response to a request, but not as blatant self promotion. Student posts are also banned, especially ones that appear to be homework assignments, but the definition of student is fuzzy (some are working engineers and part time students for instance) but FSAE is generally not red flagged, or at least they tend to not disappear.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[larrylcoyle]

Thanks Pat, FSAE allows outside services in some cases as long as the vendor is signed up with SAE and FSAE and offers that service to all teams and not just one individual. The actual request has to go through FSAE via the SAE web site though. Also, for a clarification on the Unigraphics Porting post the other day, there is a certain amount of information that would need to be exchanged that is usually covered by a NDA agreement and could not be posted openly. What I did explain though is that any information regarding Unigraphics and CNC Porting (to a point) would be openly posted so that everyone would benefit from the information. Hopefully that clears that one up as well. I did discuss that one with the "Boss Man" if you know who I mean.

I am more than willing to share what knowledge I have (to a point) that will benefit others. I have been doing this type of work too long for others not to benefit in some way, shape, form or fashion from what little I can share.

Larry

 

[SomptingGuy]

FSAE have their own forums elsewhere and they are fairly active. I don't know if they are moderated, but they do seem clear of overt advertising. Maybe it's because there's no money to be had?

- Steve

 

[mattsooty]

I must admit that I have a bit of a soft spot for the guys involved in FSAE, although I havent had much to do with it.

I wonder whether there would be any merit setting up an FSAE Forum under the Automotive Engineers part of this site? At least everything would be out in the open then?

MS

 

[SomptingGuy]

Surf here:

http://fsae.com/eve

- Steve

 

[PedroCG]

Ok, one more question, what kind of data does a grinding machine needs? i have expressed the shape of the cam in cartesian coordinates about the center of the base circle