The .tgm, and +ttool

Discussion in 'Car Modding' started by Tim Bennett, Jan 17, 2012.

  1. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Finally weather did allow some stuff done outside, so here is some photos of tire that I did cut, cost me one cutting disc, those steel plies are surprisingly hard stuff especially when rubber gets all sticky and smoky :p
    http://jtbo.pp.fi/images/album/rfactor/real_tires/

    Maybe I should still cut piece that I got out to half to find out bit more about angle of those steel parts as it is bit hard to tell what angle of those is, also difficult to cut straight. Inner layer how ever is 90 degrees, I believe as it runs directly side to side.

    I found more questions that answers though, how do I tell if Polyamide is closer to thread or rim? How do you tell difference of polyamide and polyester any way, they are plastic stuff?

    If I had to guess, I would say polyester being near to rim.

    Might try to clean cut surfaces, if that would reveal bit more stuff, maybe sandpaper would help there.

    I did not want to wreck wheel any more than it already is a wreck, so could not cut it all the way.

    Those that are not familiar with measurement too, idea is to look first how many full mm that lower 0 is, then look how many fractions to add into that reading from lower scale, when lines meet perfectly that is fraction to add.

    This tire bursted it sidewall at road construction site, rock punched trough from side so tire is not useful for anything, except maybe studying how it is made, shame really, best tire set I have ever had.

    Oh and sorry about snow, it just comes everywhere, nothing but trouble it is.
     
  2. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    I wonder here a bit about tire weight.

    Here is table of tire weights in LBS:
    http://www.miata.net/faq/tire_weights.html

    Now if I would like to make one specific tire from that table, that has weight of 20 pounds, that makes up to 9.071847kg, but ttool automatically sets weight and only way I can change that is by making changes to materials and thicknesses.

    So to get weight right, I need to play with many of those values that affect also to handling.

    Can weight then be indicator of tire being close or far?

    Probably weight is correct when all material settings are correct, so I think that might be indicator that of course is not telling which thing is wrong.



    I'm still looking table that would tell material parameters for nylon, polyester, steel and so on, I would imagine ISI should have such as they have picked values for rTrainer for example, it would be silly to hunt down those values every time making a tire.
     
  3. Domi

    Domi Registered

    Joined:
    Oct 11, 2011
    Messages:
    747
    Likes Received:
    44
    Check this line on the [QuasiStaticAnalysis] (tgm file): TotalMass=xxxx

    I think the tool gets the weight from that line.
     
  4. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    I think that tool writes weight there, but it probably does not read it from there, at least I think that I changed it, saved document, loaded it into ttool and weight was still the same?

    Changed total mass and ring mass, but either of those seemed not to change weight that is in ttool, either I made errors or it is not using those, chances are 60/50 that I made error (finnish humour) :D

    Anyway, I did read from tutorial that:
    TotalMass=10.558957270550053 // total tyre masses and inertia's as calculated by ttool upon saving a TGM file.

    It also is something that leads me suspect that we can't input mass directly, or has someone had success in that? My ttool is calculating at the moment, but might test later when calculations are complete, test 28/144 currently and it has been going for couple of hours...
     
  5. Domi

    Domi Registered

    Joined:
    Oct 11, 2011
    Messages:
    747
    Likes Received:
    44
    Hmm, I checked and you are right..
     
  6. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Ok, that is good, now there is one more thing we know, eventually we know it all bit by bit :)
     
  7. Domi

    Domi Registered

    Joined:
    Oct 11, 2011
    Messages:
    747
    Likes Received:
    44
    And by the way, the weight of the front and rear tires from the F-Renault 3.5 are correct, they match the weight of the real tires.
     
  8. patryksok.

    patryksok. Registered

    Joined:
    Oct 4, 2010
    Messages:
    130
    Likes Received:
    1
    Yes, if you want real tire weight you need build tire that have real thickness, real number of plies and materials on plies :) Imo, its realy good, when actual tire model calculate this thing due to materials and geometry. Tire model in rF2 is the most awsome I ever saw in racing sims, things we can do with tire and the potential of tire model is huge :D But if you want do good tires you need basic engineering knowledge ;) Thank god I'm mechanical engineering student ;P


    My tire have 360tests and after 6.5h ttool was done only 92 tests :p
     
  9. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    You have bit more belts then I guess ;)

    As you mentioned being engineering student, I would imagine you might have books that show modulus and that poisson ratio of different materials, would it be possible for you to dig them up and post here for those most common materials used in tire, nylon, polyester, kevlar, steel and rubber, if it states different shore values to have different numbers those would help also a lot of whole community I believe :p

    It is bit of work of course, but I think it is also big help to most of guys, especially when we set all such things to rF2 wiki it will smooth things quite a bit and increase average quality of mods compared to rF1, at least that is what my old brain tells me :D

    Another thing what I'm thinking is to get sidewalls of tires spied, there probably is mention of how many belts and what material in every tire, so would be neat to have database of those, where we would write up type of tire, size, age, just to get nice references, so after year or two we could have something from different kind of tires, there is chance someone will need such information at sometime in future.

    I have quite few tires at warehouse, so I might spy those and put it up here then, one of tire types I have, is already in link I posted earlier in photos I took it says pretty much all data there was at side of tire.
     
  10. patryksok.

    patryksok. Registered

    Joined:
    Oct 4, 2010
    Messages:
    130
    Likes Received:
    1
    I working now on tires that should be Michelin Pilot Sport 3 and PS2(when I finish work on tires I will maybe do pack with them...) i know what materials are used in carcass, but still I dont know what rubber type is "BulkMaterial" and "TreadMaterial", what are plies angles, how the plies are layed in carcass, in what "NODE" specific ply should be or not to be.

    Young modulus, Possion Ratio, Destiny, Thermal Conductivity and Specific Heat for the materials you can find even on Wiki ;) or here http://www.matbase.com/index.php or in book, industrial documents etc.

    Polyamid used in tires are mostly Nylon6-6, most common "rubbers" is Styrene-butadiene elastomers, probably with natural rubber cause Styrene-butadiene elastomers are poor in heat up build. "Solution SBR with high vinyl and styrene levels is used in high performance tire treads to improve wet traction."


    Polymer Data Handbook.
     
  11. Kev

    Kev Registered

    Joined:
    Oct 28, 2010
    Messages:
    156
    Likes Received:
    0
    Working with ttool now and i'm disappointed really. It is single-threaded. My 4-core-with-HT CPU is sleeping while ttool works hard. :(
    For such heavy calculations there should be some parallelism made.
     
  12. patryksok.

    patryksok. Registered

    Joined:
    Oct 4, 2010
    Messages:
    130
    Likes Received:
    1
    Thats probably because you cant calculate one tire physics in pararell. Because every physics process has function in time, and what would be when 1 core will calculate half tire, and other calculate other half, they need to be synchronized and will loose time or stability, probably.
     
  13. Kev

    Kev Registered

    Joined:
    Oct 28, 2010
    Messages:
    156
    Likes Received:
    0
    That's probably right (which i suspect is wrong actually) but automated testing does many independent tests. They CAN be runned in parallel threads.
    As prefessional programmer i see that it's not a big problem to run 4 tests simultaneously in 4 threads. Hope any ISI programmer take a look here and think about paralleling this tasks in ttool.
     
  14. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Tire that I cut up, there was one layer inside that runs from edge to edge, then two steel belts that run along the tread, most likely along the tread, then one layer that I have not yet been able to tell which way it goes and also I don't know if polyamide or polyester is one that is inside most layer that runs from side to side, that is 90 degrees according to thread.

    It is perhaps polyester, it looks like bit of fabric type, but hard to say.

    Then I have seen and read that radials often utilize belts under thread that run along the thread, so that it is 0 degrees and belts at sidewalls often are 90 degree angle.

    There is still lot of questions, but I think it is safe to put belts that way for PS2 and 3?

    Thread width is less than tire width and most of the belts are only under the thread, so it is nodes that are corners of tire that are start and end of most of belts, probably also no steel belt is at sidewall, so that should help to decide under which node you can put which belt.
    Also I can see from illustration that at corners there are different type of rubber, painted red in shoulder area in illustration, but which way it is different, maybe harder compound, I think that should be logical choice ?
    Thread material is material from outer limit of tire towards inside of tire by amount you specify thread depth, bulk material is all the rest - belts you specify, that is at least my current understanding, but I can't say that I would be 100% correct at this stage, this is just how I understood it from tutorial document and comments in rTrainer TGM file.

    In tire that I did cut, sidewall had 1 belt, but now I'm starting to think that sidewall belt perhaps do run also at under the thread, that is what I see in many illustrations at least, which would be 1 + 4 belts at under the tread and not just 4 and I just before was under impression of it being only at sidewall. Something to keep in mind for sure.

    I'm still trying to learn lot of stuff so I think aloud here, feel free to point to correct direction if I make mistakes.

    I did do google images search of pilot sport belts and found these, see if they are any help to your project:
    http://www.michelin.co.uk/content/img/tyres/over_the_edge_441x119.jpg
    http://michelinman.com/mediabin/App... Images/Tire Highlights Pilot Super Sport.jpg
    Second one does show angles in belts, but is it just illustration or maybe enough to be some use?

    It says super sport on many place, might be different from what you are looking, but as that is also clearly one way to construct tire, might be useful for some?

    Here is anyway text that would help you if you could find similar pic or info from pilot sport tires, I know, it is bit lot to except finding enough good pic from side of tire to be able to read those or anyone putting that info on web, but that is why I thought that everyone could hunt tires and take pics or notes and post up to make database of those, for some after some time it might be good thing.
    http://jtbo.pp.fi/images/rfactor/real_tires/belts.jpg

    Thx from those material tips, any bit can be helpful, no matter how clear it might be for someone, for someone else it might not be so clear at all :p
    I try to write up every silly thing so as I forget, I can read tomorrow how I though it yesterday :D

    Also more uncertain I'm from something, more I babble, sorry about that, but it really can be helpful sometimes ;)

    If anyone can confirm or deny my thoughts please do so, post what you think, there I learn, maybe someone else :p

    Kev, oh yes, it would be nice to have more cores computing, maybe in update or maybe it is such computing that is difficult to thread efficiently, anyway we can wish and hope :)

    edit: patrysok. I think this gives some ideas too about SP3 tire:
    http://www.synthesevideo.com/en/productions/technical-simulation/michelin-pilot-super-sport.html
     
    Last edited by a moderator: Feb 14, 2012
  15. Domi

    Domi Registered

    Joined:
    Oct 11, 2011
    Messages:
    747
    Likes Received:
    44
    Dunno if it helps... http://en.wikipedia.org/wiki/Radial_tire
     
  16. patryksok.

    patryksok. Registered

    Joined:
    Oct 4, 2010
    Messages:
    130
    Likes Received:
    1
    If you have info that tread have 4 plies one of the material if the same as it used on sidewall :p So it isnt 4+1, just 4 :p But its only my interperetation, I dont have real tire to find out that.


    I know how many and what plies Pilot Sport 3 and PS2 has :) I found those pictures :p

    No :p Because I know bias-ply and radial tire construction, we now modeling specific tire models like Pilot Sport Cup or something, and now we dont know how this specific tire is constructed.
     
  17. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Your interpretation sounds good, that would make it also easy to know that in my case, Polyester is innermost layer/belt, which would then make it quite easy to spot others from what I have managed to find out, thx :D


    Pilot sport cup can be very different from SP2 and SP3, it is DOT racing tyre and might have nothing same with SP2 and SP3 than a name, so one needs to be careful not mixing those up. It seem to have nylon layer similar to SP3 and two steel cords however, so maybe it has some things in common, but I also did read from tire rack that it is not same nylon layer, but stronger, if I understood correctly. Then again how far or close one can end up anyway without perfect specs?
     
  18. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Hmm, I'm now wondering about these two, what might be best way to choose correct values for these?
    TemporaryBristleSpring=(24000, 14800, 31500) // Lat/Vert/Long
    TemporaryBristleDamper=(0.55, 0.3, 0.55)

    Tutorial says this:
    TemporaryBristleSpring = Bristle spring rate for (Lat, Vert, Long). Temporary until contact patch model finalised. Once a TGM is generated through the run automated tests, this effects the stiffness of the bristles in the real time model (which are in contact with the road). This should be adjusted so that the real time model approximates the stiffness as seen in the QSA model. For the vertical component this can be done relatively simply by adjusting this value until the load for a given deflection in the real time section matches that obtained in the QSA section. Keep in mind at present, the tyre should be softer in real time than QSA, due to computational reasons. In other words, you may see it fit to use the bristle spring rate as a compensation for this effect.

    My QSA Model section says Y stiffness being 3090021.xxxxx so do I really set Vert tempbristlespring to that high value? It seems rather high to me, especially as it originally is so much less, should this level of spring be something in range or is it out of the whack meaning there is major issue with my built tire?

    With my earlier made tires value is over 2 million also and with those only thing changed is tire size (changed nodes) and thread depth, rest is same as original rTrainer tires, well optimal tire pressure was also set to 200 instead of 150, but other than that those earlier ones should be almost like rTrainer tires.

    But if rTrainer uses values of only 14800 for Vert(y-axis) I become very skeptical about my readings, it is huge difference. What kind of values others have got and what kind of values have you put to tempbristlespring?

    Then also come question of tempbristledamper, should original ratio between spring and damper be maintained or is there any better way determining appropriate damper values?
     
  19. patryksok.

    patryksok. Registered

    Joined:
    Oct 4, 2010
    Messages:
    130
    Likes Received:
    1
    Very alpha on materials :p
     
    Last edited by a moderator: Feb 17, 2012
  20. jtbo

    jtbo Registered

    Joined:
    Jan 12, 2012
    Messages:
    1,668
    Likes Received:
    48
    Thx from those, much easier when there is some starting point :D

    I imagine that when tire has softer Shore reading, it can be achieved by reducing bulk and thread material modulus and maybe increasing Possion ratio? It might be even possible to workout which kind of material parameters corresponds to which Shore reading as Shore reading is just softness of rubber compound.

    edit: I found some place that might be useful for some at some stage, here is search of Nylon, but of course anything can be searched, problems might come from vast range of products and how different 'blends' can be with very different properties, but at least one can get ranges which to stay with from there:
    http://www.matweb.com/search/QuickText.aspx?SearchText=Nylon 66
     
    Last edited by a moderator: Feb 15, 2012

Share This Page