LocostUSA.com

Learning how to build Lotus Seven replicas...together!
It is currently March 28, 2024, 4:49 am

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 424 posts ]  Go to page Previous  1 ... 17, 18, 19, 20, 21, 22, 23 ... 29  Next
Author Message
PostPosted: January 26, 2013, 11:38 am 
Offline
Always Moore!
User avatar

Joined: November 9, 2007, 3:40 pm
Posts: 4074
Location: Pittsburgh, PA
RbBugBitMe wrote:
Fixing 3 corners and loading 1 can be done but you have no basis to judge from so 4,000ft-lbs/deg for that won't be comparable because most everyone else with those numbers only fixes 2 points of the chassis. Fixing 3 should require higher stiffness values than 4000.


Unless I am missing something, I do not see what fixing three points will tell you if you are trying to maximize the torsional stiffness of the structure - it isn't a realistic simulation of how the chassis is actually loaded when the torsional stiffness matters.

BTW you pointing out my mistake* significantly increased my chassis's theoretical stiffness to around 10,000 ft-lbs/deg. Thanks. ;)

*As an aside - the awful professor I had for that statics class actually taught it wrong and I'm ashamed to admit that I mix up the her incorrect way and the right way. If I ever get a tattoo that equation will be a candidate. Apparently when you're taught something incorrectly it somehow gets etched into your head no matter how intuitive it is. :BH:

_________________
-Andrew
Build Log
Youtube


Top
 Profile  
Reply with quote  
PostPosted: October 23, 2013, 6:26 am 
Offline

Joined: October 22, 2013, 11:47 pm
Posts: 5
interesting subject and one that is dear to me, I have just finished building a single seater and had a good friend on the ozzclubbies forum where I am a member, do an FEA on the chassis, and we got a figure a lot lower than I was hoping for , but never the less a reasonable one, then about 3 weeks ago I put it on a test bed and got some real world figures, and they funny enough were just a bit higher than what I was aiming for, so it turned out the FEA was about 20% wrong in the wrong direction, and I must add that the person who did this for me works for a company that charges around 6-7 grand to have this done, so there not a mickey mouse outfit by any means,
In summary, I think FEA is a good guide, but I would still put my trust in a proper beam and torsion test, and if you can get your chassis around the 4000nm/pdeg your doing ok

_________________
HP = how fast you hit the wall
Torque = how far the wall moves when you hit it.

if you already no the answer,
why are you asking me the question.


Top
 Profile  
Reply with quote  
PostPosted: October 23, 2013, 7:47 pm 
Offline
Toyotaphobe
User avatar

Joined: April 5, 2008, 2:25 am
Posts: 4829
Location: Fort Worth, Texas
Glad to see you over here Kaspa. I'm Texan over there.

Now if you'd just bring your car along for test drives while you're over here.

_________________
mobilito ergo sum
I drive therefore I am

I can explain it to you,
but I can't understand it for you.


Top
 Profile  
Reply with quote  
PostPosted: October 23, 2013, 8:40 pm 
Offline
The voice of reason
User avatar

Joined: January 10, 2008, 4:47 pm
Posts: 7652
Location: Massachusetts
Quote:
In summary, I think FEA is a good guide, but I would still put my trust in a proper beam and torsion test, and if you can get your chassis around the 4000nm/pdeg your doing ok


Real world testing is the best, but the computer modeling is a great way to work towards getting to the real world unless you want a mountain of scrap out behind your shop!

One error I found a couple of times was not having all the tubes join at a node properly. It's the equivelant of forgetting to weld a joint and that would affect your real world stiffness too. The way I found it was by watching the chassis flex in the animation and noticing tubes pulling away from each other.

It would happen when deciding to add a diagonal. You add the node and put in the diagonal, but you have to remember to remove the original tube and then put two tubes in it's place. Over time as you keep trying different things eventually you can forget that.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.


Top
 Profile  
Reply with quote  
PostPosted: October 24, 2013, 2:32 am 
Offline

Joined: October 22, 2013, 11:47 pm
Posts: 5
Quote:
One error I found a couple of times was not having all the tubes join at a node properly. It's the equivelant of forgetting to weld a joint and that would affect your real world stiffness too. The way I found it was by watching the chassis flex in the animation and noticing tubes pulling away from each other.


that's an extremely good point , and one that I find on almost every second car I look at , but for some unknown reason they wont change,, also adding a small brace to certain sectors will add a lot of strength.

cheers Kaspa

_________________
HP = how fast you hit the wall
Torque = how far the wall moves when you hit it.

if you already no the answer,
why are you asking me the question.


Top
 Profile  
Reply with quote  
PostPosted: November 10, 2013, 12:55 pm 
Offline
The voice of reason
User avatar

Joined: January 10, 2008, 4:47 pm
Posts: 7652
Location: Massachusetts
Andrew posted this model in Grape of the Kurtis Kraft 500 chassis. This is the first take on it. I am also going to update the first post in this thread with a table of contents of where models are posted here now that this thread is become many pages long.

Image


You do not have the required permissions to view the files attached to this post.

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.


Top
 Profile  
Reply with quote  
PostPosted: December 21, 2013, 7:54 pm 
Offline

Joined: December 21, 2013, 7:43 pm
Posts: 1
Contributing to the body of knowledge on chassis stiffness and analysis, I've recently uploaded a report analysing a Westfield chassis. The study includes torsional and bending stiffness, but focusses on torsional stiffness. The chassis stiffness workshop tested and used to validate a computer model to explore options for modifying the chassis. A summary of about 30 modifications including their effect on weight and stiffness is presented. Drawings and data files are included as attachments.

The report can be viewed downloaded from the following link
http://www.scribd.com/doc/192898005/Veh ... Sports-Car

[url]http://www.scribd.com/doc/192898005/Vehicle-Structures-Development-of-the-Sports-Car-Chassis-and-Stiffness-Analysis-of-the-Westfield-Sports-Car]Westfield Chassis Analysis Report[/url]


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 8:36 am 
Offline
The voice of reason
User avatar

Joined: January 10, 2008, 4:47 pm
Posts: 7652
Location: Massachusetts
Thanks for posting this Ganhaar! I'll start reading…

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 12:05 pm 
Offline
User avatar

Joined: October 24, 2008, 2:13 pm
Posts: 5326
Location: Carlsbad, California, USA
That's almost a complete educational piece. I skimmed it and then downloaded it thereafter. Thank you for posting it, Ganhaar.

As a funny aside, I love how the old chassis examples and illustrations from Costin and Phipps keep reappearing in papers. Some day my original copy of their book may be worth a lot of money.

I've not heard of PAFEC finite element software before, which is not in anyway meant to imply a negative. I love the fact that the theoretical prediction of the software was validated in the lab. The reported accuracy to 1% in torsional stiffness prediction by the software is most excellent. The 11% accuracy in bending was attributable to the use of flexible wooden supports in the lab verifications tests according to the authors, which they acknowledge as an oversight.

I'll read it in detail in the near future. It's very good stuff too with lots of historical data. Be prepared to convert the Newton-meters to Pound-feet and kilograms to pounds, etc. Google will do that for you if you enter it as a search phrase, so no whining. :lol:

Cheers,

_________________
Damn! That front slip angle is way too large and the Ackerman is just a muddle.

Build Log: viewtopic.php?f=35&t=5886


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 1:00 pm 
Offline

Joined: October 19, 2009, 9:36 pm
Posts: 2199
Location: meadview arizona
well, like most people without a background knowledge in chassis design the tests look very scientific.

thats where i differ in my opinion, the bending tests, they used the wrong points for the wooden blocks, they used the wrong points to place the loads, they did not take into account any bending of the chassis at right angles to the center line and assumed that the points chosen for measurement were in fact the lowest points on the chassis when loaded.

in fact the position of the wooden blocks is a critical mistake, if they had done the bending test at the correct points, supporting the chassis at the suspension pickup points, the ends of the chassis would have shown a negative load.

in the torsional test they did not account for deformation at varying points down the chassis and appeared to treat the chassis as a solid object, this may be helpful to achieving the results they wanted but is of little use in determining where the chassis is deforming and in which direction at different points.

what would have been more useful would have been to fix the rear suspension points then twist the chassis and measure at varying points like the back of the cockpit, the front of the cockpit, the front of the engine bay and then overall at the front suspension mounts, this would have shown where the most twisting occured, thus showing where additional bracing is required and this should have been done in two planes as the chassis may not rotate about its imaginary axis but bend as opposed to twist.

a locost chassis forward of the rear cockpit bulkhead is more like two structures joined with the top of the tunnel and reverts to a single structure forward of the footwells, thus there is far more going on here than a simple beam twisting about its axis.

just my opinion.

_________________
this story shall the good man teach his son,
and chrispin chrispian shall ne'er go by,
from this day to the end of the world.
but we in it shall be remembered.


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 3:06 pm 
Offline

Joined: October 19, 2009, 9:36 pm
Posts: 2199
Location: meadview arizona
if you look at the chassis as a spring, well actually a series of springs of varying tension joined together.

each part of the chassis has its own perpencity to be deformed or loaded independant of the other parts and still retain its abillity to recover.

if a twisting moment is placed upon the chassis, it will deform the lightest part first, to a point where the tension in that part, or spring, reaches a level that the next heaviest part begins to deform, at which time both will deform, the later at a slower rate, until the next heaviest part and on and on until recovery is lost in the lightest part, thus causing a failure in that part but only in that part, the other parts still have the abillity to recover.

this is at the heart of a stiff chassis, not the overall stiffness as measured at the ends, we are looking at the weak link in the chain, not the potential strength of an average of random sized chain links.

this would let us root out the weak sections of the chassis, which in turn, we can break down yet again to discover the weak part of that section, if we do not do this, then our only way to stiffen the chassis is everywhere, adding tubes infinitem.

we must twist each section individually, then in pairs as they assemble and twist again until each section is as stiff as the other but no more than any of the others, then on assembly an overall chassis, twist can be measured accurately in the knowledge that there are no weak links in the chain and any twist is spread evenly throughout the chassis. without the need for tubes that do nothing.

i must be noted that some sections of the frame are longer and wider/taller than others so measurements should be taken at equal increments, say 6", down the whole chassis as a long section will twist more than a short section, this will give a figure of twist per 6" for a given torque applied rather than one section compaired to another.

when you have reached a twist figure that can not be measured at 6" increments, then go to 12" increments.

again be mindfull that twist is actually occuring within a section and not deformation of each "bulkhead" which should remain true but rotate about a common axis at all times, if this axis deviates then the chassis is not twisting but buckling and any and all information about twist has gone out the proverbial window because the chassis is no longer "square" in relation to the suspension pickup points.

note that there are many factors which have an effect on the stiffness of a chassis that are not in the bare chassis, be especially mindfull about engine and transmission, driver & passenger loads and their perpencity to stay moving in one direction, motor mounts are in essents springs with a limited compliance, likewise seat foam, once that compliance is reached, they will be imparting a bending load on the chassis in a turn, if i doubt, stand your chassis on its side and put the engine and trans on the top, then stand on it and see if it deflects.

_________________
this story shall the good man teach his son,
and chrispin chrispian shall ne'er go by,
from this day to the end of the world.
but we in it shall be remembered.


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 4:22 pm 
Offline
User avatar

Joined: February 20, 2009, 2:27 pm
Posts: 531
Location: Reno, Nv
John
I think the testing and results were better then you indicate, though I still haven't read the whole article it appears the focus was to determine if the FEA program was providing accurate results and if they restrained the chassis at the same points as the model the results look pretty good to me. As far as the twisting of the chassis they used the suspension point to restrain and measured in enough areas to determine that the engine bay was the weak point so the chassis was divided up somewhat.
I found the article of interest and usefull.

_________________
John - Slow and Steady. . . Well slow anyway
Build Loghttp://www.locostusa.com/forums/viewtopic.php?f=35&t=6245


Top
 Profile  
Reply with quote  
PostPosted: December 23, 2013, 7:57 pm 
Offline

Joined: October 19, 2009, 9:36 pm
Posts: 2199
Location: meadview arizona
any knowledge is good knowledge!!!!!!

its being able to apply that knowledge in a practical way that requires experience in the field.

in general terms, the article was informative but i just needed to point out where its limitations lay and what needed to be considered in addition to the article as opposed to following blindly.


as i told my wife when she was complaining about a young supervisor with a degree, just out of college, a degree doesn't make you an expert, nor does 30 years of experience if it was the same experience every year, a prooven track record does!!!!!

i remember asking an engineering student "how many examples of load over crossectional area he could come up with" all he could manage was "stilletto heels"

_________________
this story shall the good man teach his son,
and chrispin chrispian shall ne'er go by,
from this day to the end of the world.
but we in it shall be remembered.


Top
 Profile  
Reply with quote  
PostPosted: December 24, 2013, 12:13 am 
Offline
Toyotaphobe
User avatar

Joined: April 5, 2008, 2:25 am
Posts: 4829
Location: Fort Worth, Texas
john hennessy wrote:

i remember asking an engineering student "how many examples of load over crossectional area he could come up with" all he could manage was "stilletto heels"


At his age that was probably something that was on his mind a lot.

What do I mean "at his age"? at any age!

_________________
mobilito ergo sum
I drive therefore I am

I can explain it to you,
but I can't understand it for you.


Top
 Profile  
Reply with quote  
PostPosted: December 24, 2013, 1:06 am 
Offline
The voice of reason
User avatar

Joined: January 10, 2008, 4:47 pm
Posts: 7652
Location: Massachusetts
I read thru most of the paper so far. I haven't had time to carefully look at the diagrams to see all the actual changes. I think it's quite a good work for a collage student. Computing resources where much less available in 1992. Considering the 20+ years that have passed that would be about a 50,000 time improvement in computing resources.

At the end of the paper there are comparisons with other frame types that are interesting.

John, I disagree with your basic assessments of this paper. For instance you say:

Quote:
in the torsional test they did not account for deformation at varying points down the chassis and appeared to treat the chassis as a solid object, this may be helpful to achieving the results they wanted but is of little use in determining where the chassis is deforming and in which direction at different points.


The point of this paper is that the testing was done with computer models. The model would generate displacements of every node in the chassis in 3 dimensions to a good number of decimal places. So there is no reason to think this student treated the chassis as a solid object.

The computer modeling makes it easy to study how a frame really reacts. In much more detail than you can get just with dial gauges. Dial gauges let you see how the frame moves but don't tell you what the loads on the tubes are. You have to infer that from experience or sit down and do potentially a lot of math. It's not realistic to do it by hand really.

I also think breaking the design down into little boxes can be useful, but by no means should it be the only or basic way you approach the design. I remember the first time I changed a tube at the very back of my frame and it caused the diagonals across the front face of the nose to light up. It was exciting but also forced me to try to also think of the whole frame as a system. The feature that allows the software to animate the model and show it flexing back and forth with greatly exaggerated motion is also a big help. Seeing how the bullheads warp showed me how to build more stiffness into the frame.

All in all I think it was a commendable effort from a student…

_________________
Marcus Barrow - Car9 an open design community supported sports car for home builders!
SketchUp collection for LocostUSA: "Dream it, Build it, Drive it!"
Car9 Roadster information - models, drawings, resources etc.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 424 posts ]  Go to page Previous  1 ... 17, 18, 19, 20, 21, 22, 23 ... 29  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 9 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
POWERED_BY