Sign in or 
Share this
Suspension Group
You can edit and add more pages with descriptions, diagrams etc. or ask me questions via the message box below. You can even attach files providing they are not too large. Please use this as your CODEX for investigations and insights into your project.
MASSIMO
TAYBO
ANTONIO
GERARD
JAVIER
VICTOR
RUI
GABRIEL
MIGUEL
SERGIO
JOSEPH
This way, all the students can benefit. It enables everyone to learn from each other, creating a real educational experience.
My observations and comments are made on a page by page basis related to the “RCC_19_100711_General” document as placed into Dropbox.
I must first say that this report ranks as one of the top reports that I have had the pleasure of reading from METCA students.
The group has conducted themselves very professionally. Not only that, they have worked as a team. The concept of a team dynamic, working towards a common aim of understanding how racing cars work and ultimately making “cars go faster” is just as important as individual effort.
I hope you will have time before my visit to make amendments to your project based upon my feedback. I am going to make constructive criticisms. That is, I will point you in a particular direction so that you can draw your own conclusions. This encourages discussion and debate.
Page 4.
I very much like your planning and management system. It aids communication and delegation.
Page 5.
You are right to establish your frame of reference right from the start. Unfortunately your statement that the “CAD and kinematic analysis will use the same system is confusing.” I can see why you have attempted to do this, in fact I applaud your lateral thinking behind it. However, we try to prevent confusion by adopting conventions. Historically, the drawing office convention is Cartesian based, with a two dimensional drawing plane, as if you are looking at the object from outside. Up is then positive from the origin, down is negative. Right is positive from the origin and left is negative. Now we have three dimensional CAD, but the Cartesian convention is maintained. Therefore, if you were drawing the car looking at it from the front the convention would be as follows: This has the X axis as positive going through the drawing plane towards the rear of the car. The origin will be the O-Line bottom of chassis and on the centre line of the car. These are (Vehicle Coordinates) with the chassis on a horizontal plane. Everything to the right on the drawing plane Y axis is positive. Everything above the O-Line Z axis on the drawing plane is positive. However when we drive a vehicle we are on the inside looking out. The car is now sitting on the ground (Earth Coordinates) not CAD (Vehicle Coordinates). It is then normal to have the X axis facing forwards in the direction of travel. Two systems prevail being SAE and ISO. In the SAE system the Z axis is positive downwards, this is very confusing for most people other than the Americans. In Europe the ISO system is the normal convention. The ISO vehicle axis system (ISO 8855) uses the right hand rule ; X forwards, Y left and Z up, but as long as you state your system as clearly as possible then no problems should occur. Wheels and tyres have their own axis systems, which then adds to the confusion. The upshot of this all, is that, it’s not the convention that’s important, as long as you make it clear what convention you are using.
As you can see conventions and conversion between coordinate systems is the cause of an unimaginable amount of “****-ups” in all fields of scientific study. I hope this helps, if not, then join the majority of people that live in blissful ignorance. Sometimes I wish that I could join that group too. Life would be so much easier!
Please take a look at the chapters:
Conventions and Vehicle Axis Coordinate Systems on this wiki.
Page 6.
I can see from the drawings that a vehicle based, right hand rule convention has been applied with the X axis pointing forwards. As long as I know the convention you are using and appropriate conversions are made when the kinematic (Earth Coordinates) analysis is made, then no problems should occur.
Page 7.
You have attempted to do a kinematic analysis by warping the ground. This is quite a bold concept, and one that I often use myself. It has a number of advantages. However, it is very easy to run into problems in the kinematic analysis because you are not modelling reality correctly. What you have done is to create an abstract reality which will require careful manipulation before any kinematic analysis can take place.
I would like to hear the reasons why you have chosen this rather unconventional but interesting approach to analysing the suspension geometry.
Page 21.
It is important to list the assumptions used and limitations of your models. In most cases you have done this. As far as longitudinal acceleration is concerned under braking, what assumptions as to braking proportion front to rear were made? Under braking, the percentage anti-dive and anti-rise must be factored by the braking proportion to find weight transfers acting directly through the links.
Please refer to my comments on load transfers from page 27 onwards.
Page 25.
The aero maps are difficult to read. Three dimensional RSM representations as well as contoured two dimensional plots would help everyone to assimilate the information better.
Page 27 and 28.
Conceptual diagrams like the one you have shown certainly help everyone to visualise what parameters contribute towards the load transfer. However the diagram must agree with the later equations. State what the L terms in the equations refer to. On one of the diagrams you have shown the tyres as springs, and in the other they are not included. Should the tyre spring rates be included in the equations? There are no right or wrong ways to calculate load transfer.
Please refer to:
Lateral Weight Transfer on this wiki for an alternative approach.
At the end of the day, we want to try and model reality as close as possible. Compare the two approaches, and the results. Do they agree? If not, why not? What assumptions have been made? What are the limitations of both models?
Please also refer to:
Force Roll Centre on this wiki for a clarification on the calculation of the actual roll axis position.
In the real world, as the car rolls due to a lateral acceleration, the tyres see the weight transfer through the suspension links (geometric) and unsprung first. This has a large effect on transient behaviour. Only when the car reaches its maximum roll angle in steady state do the link, unsprung and elastic weight transfers balance out. In reality, there must be a geometric roll axis (links), a kinematic roll axis (elastic) and an aerodynamic roll axis. Check your calculations against the total amount of LWT that must take place, which is dependent on only four factors, this is if we exclude aerodynamic and steering effects:
Then check the distribution of lateral WT using both approaches. Decide what the limitations of both models are and if the models can evolve to include second order effects such as steering geometry.
Page 29.
All the observations and comments that I have made regarding the lateral WT also apply to the longitudinal WT. An important distinction here is that in the lateral case we sometimes assume that the lateral forces from both tyres on an axle are equal. In the longitudinal case we cannot assume the longitudinal forces from the front and rear tyres are equal. Somehow we must take the braking proportion into account. This has a profound effect on the amount of anti-dive and anti-rise. In the accelerating case the amount of anti-squat must be calculated in a different way. State what the W terms in the equations refer to.
I have not finished the chapter on longitudinal weight transfer on this wiki. I have left sections out on purpose. I want you to come up with a model that better represents the real world situation. Compare your diagram with your equations. Do the two agree? Is the pitch centre that is shown in the diagram, the geometric or the kinematic? Should we include the tyre stiffness? Are the geometric and kinematic pitch centres in conflict in any way? Will there be a shift in aerodynamic balance as the car pitches?
Pitch sensitivity is very complex subject and misconceptions abound. If you try to oversimplify the calculations of longitudinal WT it can have very dangerous implications. Remind me too show you two videos on the subject of pitch sensitivity and what can go seriously wrong during my next visit.
To start you off, look at the diagrams below. They show some basic concepts. Try to compare these diagrams, equations and results with yours. Do they agree? If not, why not? What assumptions have been made? Come up with revised equations and validate these results with the WS car. Compare your results for the anti’s with those stated in the manual.
The next diagram shows the geometric condition of 100% Anti-dive and 100% anti-lift. How do we get the proportoin of anti's that we have in the WS car?
As in the lateral case where you can find the force roll centre, finding the force pitch centre is fundamental to anti concepts.
Read again the following chapters:
The Design Process on this wiki
The Three Forms of Knowledge on this wiki
Base your new diagrams and equations on first principles. Do not be persuaded by my or other people’s ideas.
I am reminded of a quote from B. H. McGill
Education should prepare our minds to use its own powers of reason and conception rather than filling it with the accumulated misconceptions of the past.
Bryant H. McGill
Pages 30 to 34.
I like your response surface analysis. A three dimensional picture conveys so much more information. However, the ranges you have picked are too high. A slip ratio of +-10% and slip angles of +-10 degrees is the maximum degree of confidence that this MF-Tyre model can provide. I would like to see more RSM for camber, normal load and toe sensitivity. Maybe this could also include trend surfaces. Remember that the tyre properties and characteristics are crucial to the suspension conceptual design. Try to acquire as much information as you can about this most important component on the race car!!!
Page 35.
You cannot do better for your reference than Mike Blundell and Damian Harty.
Did you validate your findings with the geometry program link loads I gave you?
Page 42.
I do not agree with your statement: “In order to have a more stable car the roll centre should be on the ground…” Please explain. Your statement may be due to misconceptions again.
Page 43.
This page is full of misconceptions. Try to be very clear about the pitch concepts here. Under braking with brake discs at the wheels, we can employ an amount of anti-dive (front) and anti-rise (rear) to control pitch motions. Under acceleration, as we only have rear wheel drive there can only be anti-squat (rear). No anti rise (front) is possible due the wishbone geometry alone. After revising your pitch analysis equations, all this should become clear.
Pages 45 to 47.
The diagrams are not too clear as the ranges chosen are too high. I would like to see some validation of the corner weight results.
Pages 50 to 54.
This is the part of your project where I had great expectations. It’s the simulation part where we get all the answers to our “what if questions.” You have fallen into the “paralysis of analysis” trap.
Read the chapter:
The K.I.S.S. Principle on this wiki
What we ideally want is a simple model that can follow our analysis process very closely. This way optimization of conceptual design parameters can be made along the way. On page 20 you showed a G-G diagram. If you add speed to this chart you would then have a picture of the car working within its performance envelope. The logical next step would be to have a simple parametric model of the car and plot say the understeer angle within the performance envelope. This will provide a handling profile for the car at any speed, corner radius and lateral acceleration combination that the car will ever experience. An optimization of the parameters can then take place.
I will state again here a definition of what is meant by simulation.
Simulation is the imitation of some real thing or process. It entails representing certain key characteristics or behaviors of a selected physical or abstract system, thus enabling one to perceive the interactions that would not otherwise be apparent.
I look forward to reading your amendments.
Regards, Karl
MASSIMO
TAYBO
ANTONIO
GERARD
JAVIER
VICTOR
RUI
GABRIEL
MIGUEL
SERGIO
JOSEPH
White-Smoke provisional evaluation on the RCC Suspension Assignment
As stated above this will be a provisional evaluation of the project. I would like to go through each project with all the students when I am at METCA again at the end of October.This way, all the students can benefit. It enables everyone to learn from each other, creating a real educational experience.
My observations and comments are made on a page by page basis related to the “RCC_19_100711_General” document as placed into Dropbox.
I must first say that this report ranks as one of the top reports that I have had the pleasure of reading from METCA students.
The group has conducted themselves very professionally. Not only that, they have worked as a team. The concept of a team dynamic, working towards a common aim of understanding how racing cars work and ultimately making “cars go faster” is just as important as individual effort.
I hope you will have time before my visit to make amendments to your project based upon my feedback. I am going to make constructive criticisms. That is, I will point you in a particular direction so that you can draw your own conclusions. This encourages discussion and debate.
Page 4.
I very much like your planning and management system. It aids communication and delegation.
Page 5.
You are right to establish your frame of reference right from the start. Unfortunately your statement that the “CAD and kinematic analysis will use the same system is confusing.” I can see why you have attempted to do this, in fact I applaud your lateral thinking behind it. However, we try to prevent confusion by adopting conventions. Historically, the drawing office convention is Cartesian based, with a two dimensional drawing plane, as if you are looking at the object from outside. Up is then positive from the origin, down is negative. Right is positive from the origin and left is negative. Now we have three dimensional CAD, but the Cartesian convention is maintained. Therefore, if you were drawing the car looking at it from the front the convention would be as follows: This has the X axis as positive going through the drawing plane towards the rear of the car. The origin will be the O-Line bottom of chassis and on the centre line of the car. These are (Vehicle Coordinates) with the chassis on a horizontal plane. Everything to the right on the drawing plane Y axis is positive. Everything above the O-Line Z axis on the drawing plane is positive. However when we drive a vehicle we are on the inside looking out. The car is now sitting on the ground (Earth Coordinates) not CAD (Vehicle Coordinates). It is then normal to have the X axis facing forwards in the direction of travel. Two systems prevail being SAE and ISO. In the SAE system the Z axis is positive downwards, this is very confusing for most people other than the Americans. In Europe the ISO system is the normal convention. The ISO vehicle axis system (ISO 8855) uses the right hand rule ; X forwards, Y left and Z up, but as long as you state your system as clearly as possible then no problems should occur. Wheels and tyres have their own axis systems, which then adds to the confusion. The upshot of this all, is that, it’s not the convention that’s important, as long as you make it clear what convention you are using.
As you can see conventions and conversion between coordinate systems is the cause of an unimaginable amount of “****-ups” in all fields of scientific study. I hope this helps, if not, then join the majority of people that live in blissful ignorance. Sometimes I wish that I could join that group too. Life would be so much easier!
Please take a look at the chapters:
Conventions and Vehicle Axis Coordinate Systems on this wiki.
Page 6.
I can see from the drawings that a vehicle based, right hand rule convention has been applied with the X axis pointing forwards. As long as I know the convention you are using and appropriate conversions are made when the kinematic (Earth Coordinates) analysis is made, then no problems should occur.
Page 7.
You have attempted to do a kinematic analysis by warping the ground. This is quite a bold concept, and one that I often use myself. It has a number of advantages. However, it is very easy to run into problems in the kinematic analysis because you are not modelling reality correctly. What you have done is to create an abstract reality which will require careful manipulation before any kinematic analysis can take place.
I would like to hear the reasons why you have chosen this rather unconventional but interesting approach to analysing the suspension geometry.
Page 21.
It is important to list the assumptions used and limitations of your models. In most cases you have done this. As far as longitudinal acceleration is concerned under braking, what assumptions as to braking proportion front to rear were made? Under braking, the percentage anti-dive and anti-rise must be factored by the braking proportion to find weight transfers acting directly through the links.
Please refer to my comments on load transfers from page 27 onwards.
Page 25.
The aero maps are difficult to read. Three dimensional RSM representations as well as contoured two dimensional plots would help everyone to assimilate the information better.
Page 27 and 28.
Conceptual diagrams like the one you have shown certainly help everyone to visualise what parameters contribute towards the load transfer. However the diagram must agree with the later equations. State what the L terms in the equations refer to. On one of the diagrams you have shown the tyres as springs, and in the other they are not included. Should the tyre spring rates be included in the equations? There are no right or wrong ways to calculate load transfer.
Please refer to:
Lateral Weight Transfer on this wiki for an alternative approach.
At the end of the day, we want to try and model reality as close as possible. Compare the two approaches, and the results. Do they agree? If not, why not? What assumptions have been made? What are the limitations of both models?
Please also refer to:
Force Roll Centre on this wiki for a clarification on the calculation of the actual roll axis position.
In the real world, as the car rolls due to a lateral acceleration, the tyres see the weight transfer through the suspension links (geometric) and unsprung first. This has a large effect on transient behaviour. Only when the car reaches its maximum roll angle in steady state do the link, unsprung and elastic weight transfers balance out. In reality, there must be a geometric roll axis (links), a kinematic roll axis (elastic) and an aerodynamic roll axis. Check your calculations against the total amount of LWT that must take place, which is dependent on only four factors, this is if we exclude aerodynamic and steering effects:
- Lateral acceleration
- Front and rear tracks
- Height of the CofG (total)
- Mass (total)
Then check the distribution of lateral WT using both approaches. Decide what the limitations of both models are and if the models can evolve to include second order effects such as steering geometry.
Page 29.
All the observations and comments that I have made regarding the lateral WT also apply to the longitudinal WT. An important distinction here is that in the lateral case we sometimes assume that the lateral forces from both tyres on an axle are equal. In the longitudinal case we cannot assume the longitudinal forces from the front and rear tyres are equal. Somehow we must take the braking proportion into account. This has a profound effect on the amount of anti-dive and anti-rise. In the accelerating case the amount of anti-squat must be calculated in a different way. State what the W terms in the equations refer to.
I have not finished the chapter on longitudinal weight transfer on this wiki. I have left sections out on purpose. I want you to come up with a model that better represents the real world situation. Compare your diagram with your equations. Do the two agree? Is the pitch centre that is shown in the diagram, the geometric or the kinematic? Should we include the tyre stiffness? Are the geometric and kinematic pitch centres in conflict in any way? Will there be a shift in aerodynamic balance as the car pitches?
Pitch sensitivity is very complex subject and misconceptions abound. If you try to oversimplify the calculations of longitudinal WT it can have very dangerous implications. Remind me too show you two videos on the subject of pitch sensitivity and what can go seriously wrong during my next visit.
To start you off, look at the diagrams below. They show some basic concepts. Try to compare these diagrams, equations and results with yours. Do they agree? If not, why not? What assumptions have been made? Come up with revised equations and validate these results with the WS car. Compare your results for the anti’s with those stated in the manual.
The next diagram shows the geometric condition of 100% Anti-dive and 100% anti-lift. How do we get the proportoin of anti's that we have in the WS car?
As in the lateral case where you can find the force roll centre, finding the force pitch centre is fundamental to anti concepts.
Read again the following chapters:
The Design Process on this wiki
The Three Forms of Knowledge on this wiki
Base your new diagrams and equations on first principles. Do not be persuaded by my or other people’s ideas.
I am reminded of a quote from B. H. McGill
Education should prepare our minds to use its own powers of reason and conception rather than filling it with the accumulated misconceptions of the past.
Bryant H. McGill
Pages 30 to 34.
I like your response surface analysis. A three dimensional picture conveys so much more information. However, the ranges you have picked are too high. A slip ratio of +-10% and slip angles of +-10 degrees is the maximum degree of confidence that this MF-Tyre model can provide. I would like to see more RSM for camber, normal load and toe sensitivity. Maybe this could also include trend surfaces. Remember that the tyre properties and characteristics are crucial to the suspension conceptual design. Try to acquire as much information as you can about this most important component on the race car!!!
Page 35.
You cannot do better for your reference than Mike Blundell and Damian Harty.
Did you validate your findings with the geometry program link loads I gave you?
Page 42.
I do not agree with your statement: “In order to have a more stable car the roll centre should be on the ground…” Please explain. Your statement may be due to misconceptions again.
Page 43.
This page is full of misconceptions. Try to be very clear about the pitch concepts here. Under braking with brake discs at the wheels, we can employ an amount of anti-dive (front) and anti-rise (rear) to control pitch motions. Under acceleration, as we only have rear wheel drive there can only be anti-squat (rear). No anti rise (front) is possible due the wishbone geometry alone. After revising your pitch analysis equations, all this should become clear.
Pages 45 to 47.
The diagrams are not too clear as the ranges chosen are too high. I would like to see some validation of the corner weight results.
Pages 50 to 54.
This is the part of your project where I had great expectations. It’s the simulation part where we get all the answers to our “what if questions.” You have fallen into the “paralysis of analysis” trap.
Read the chapter:
The K.I.S.S. Principle on this wiki
What we ideally want is a simple model that can follow our analysis process very closely. This way optimization of conceptual design parameters can be made along the way. On page 20 you showed a G-G diagram. If you add speed to this chart you would then have a picture of the car working within its performance envelope. The logical next step would be to have a simple parametric model of the car and plot say the understeer angle within the performance envelope. This will provide a handling profile for the car at any speed, corner radius and lateral acceleration combination that the car will ever experience. An optimization of the parameters can then take place.
I will state again here a definition of what is meant by simulation.
Simulation is the imitation of some real thing or process. It entails representing certain key characteristics or behaviors of a selected physical or abstract system, thus enabling one to perceive the interactions that would not otherwise be apparent.
Conclusion
Congratulations!!! On the whole, this was a very good report. You have time before my next visit to knock this report into shape and produce something excellent.I look forward to reading your amendments.
Regards, Karl
white-smoke |
Latest page update: made by white-smoke
, Sep 17 2010, 11:28 AM EDT
(about this update
About This Update
Edited by white-smoke
view changes - complete history) |
|
Keyword tags:
None
More Info: links to this page
|
| Started By | Thread Subject | Replies | Last Post | ||
|---|---|---|---|---|---|
| Gerard_Pérez | Urban car natural frequency | 3 | Sep 11 2010, 4:50 AM EDT by white-smoke | ||
|
Thread started: Sep 7 2010, 10:16 AM EDT
Watch
Hi Karl. I am looking for the natural frequency of road cars for my master final project. Do you know any database where this value is presented for road cars? I found a database of the NHTSA which shows the moments of inertia in the 3 main axis, but nothing related to natural frequency. Particularly I am interested to know the natural frequency of a urban "small" car, i.e. a Smart.
Thank you very much! |
||||
| Gerard_Pérez | RCC: Anti-roll and anti-pitch stiffness | 5 | Aug 12 2010, 11:09 AM EDT by Gerard_Pérez | ||
|
|
Thread started: Aug 8 2010, 11:10 AM EDT
Watch
Hi Karl. I'm from the suspension group and have been doing a MATLAB program for load transfer calculation. When calculating the longitudinal load transfer, I need finally the anti-pitch stiffness to get the pitch angle, and the same for lateral WT and the roll angle with the roll stiffness.
What I realised is that the roll angle obtained was pretty low (< 0.1 deg) in every case. This leads to several questions presented here. In the Dallara WS manual, the roll stiffness of the Belleville stack is presented in "Stiffness on ground" in kgm/deg, but I don't know how they get this value from the N/mm on ground value. What I have tried is: k(Nm/deg) = k(N/mm) * Track/2 * 1/atan(1/(Track/2)), but the value I get is pretty different from the one in the manual. Then, regarding the anti-pitch stiffness, as the value is in the spring in lb/in, I converted it to N/mm, passed to stiffness on ground dividing by MR^2, and using the same operation described before I get a value in Nm/deg, but actually I DON'T trust it anymore as the previous of roll is not matching what Dallara says. Finally, the stiffness are calculated like this: - F Antipitch stiffness: Monoshock spring on ground - R Antipitch stiffness: 2*rear spring elastic constant, on ground - F Antiroll stiffness: Belleville stack on ground - R Antiroll stiffness: Antiroll bar stiffness on ground + R Antipitch stiffness What is going wrong with the calculations? How Dallara obtains the kgm/deg value from N/mm? Are the "anti" stiffness calculated properly this way? Vielen dank |
||||
| white-smoke | DropBox now active!!! | 0 | Jun 1 2010, 1:57 AM EDT by white-smoke | ||
|
Thread started: Jun 1 2010, 1:57 AM EDT
Watch
A shared METCA folder has been set up in DropBox for you. This has been done for several reasons: 1. It is a shared folder accessible only by the METCA students. 2. You can work on project files from several different computers and the files will be synchronised automatically. 3. All folders and files are in one place and can be accessed from any computer via the DropBox web site.
Please do not alter any of the files in folders other than in your study groups, as these are master copies. You can however make your own copies of these files. |
||||
Showing 3 of 3 threads for this page
