The online racing simulator
LFS Gear Ratio Calculator 2½ [deceased]
(76 posts, started )
Good work again Bob
One thing I've been meaning to ask you about since I 'discovered' it a few days ago is the effect of drag on the load distribution. I was trying to add downforce to a weight distribution spreadsheet I made and tried to calculate the locations of the three downforce points on the FXO GTR.
It's a standard three unknowns requiring three simultaneous equation problem so I used F1PerfView to give me the wheel loads for three levels of downforce all at 125mph. The locations my calculations gave me were completely wrong so I investigated...

I was taking moments about the rear axle to calculate the load distribution which means that ALL forces which have a moment about the rear axle must be included. This includes drag! I was just wondering whether your program takes drag into account when calculating the load distribution?

Oh, and I spotted a mistake in the latest version of the Advanced Setup Guide. You removed all the stuff about downforce affecting suspension frequencies from the 'Suspension' section but not the 'Downforce' section (it's on page 26).
Quote from Stuff :Also Gizz, and Bob, don't forget about the essential .dll file, the VB 6 runtime. ...
M$ site.

Ah cheers, have added a link to that on the download page now. Well done from refraining from mentioning APIs.

Quote from StewartFisher :Good work again Bob
I was taking moments about the rear axle to calculate the load distribution which means that ALL forces which have a moment about the rear axle must be included. This includes drag! I was just wondering whether your program takes drag into account when calculating the load distribution?

Do you mean the weight transfer caused by aero forces not acting at the height of CoG? If so, then no (not even for non-downforce cars). At the moment I'm working with driving force and rolling resistance working at ground level, and drag sufficiently above the CoG height to remove any forward weight transfer caused by rolling resistance but not enough to create any rearward transfer. So basically, only accleration or braking causes weight transfer, not deceleration caused by coasting.
I would like to include these effects but have any data to work with. Even my wing/undertray x axis locations are guesstimates.

Quote from StewartFisher :
Oh, and I spotted a mistake in the latest version of the Advanced Setup Guide. You removed all the stuff about downforce affecting suspension frequencies from the 'Suspension' section but not the 'Downforce' section (it's on page 26).

:doh: ta
Quote from Bob Smith :Do you mean the weight transfer caused by aero forces not acting at the height of CoG? If so, then no (not even for non-downforce cars). At the moment I'm working with driving force and rolling resistance working at ground level, and drag sufficiently above the CoG height to remove any forward weight transfer caused by rolling resistance but not enough to create any reward transfer. So basically, only accleration or braking causes weight transfer, not deceleration caused by coasting.

Sort of, but surely it doesn't matter where the aero forces act relative to the CoG? I haven't thought about it in great detail but surely any force which does not act through one of the axles must be taken into account when calculating load distribution?

Quote :I would like to include these effects but have any data to work with. Even my wing/undertray x axis locations are guesstimates.

I would have tried it myself but 6 unknowns would require 6 sets of RAF data and 6 simultaneous equations...no thanks!
Quote from StewartFisher :Sort of, but surely it doesn't matter where the aero forces act relative to the CoG? I haven't thought about it in great detail but surely any force which does not act through one of the axles must be taken into account when calculating load distribution?

It's the CoG height that matters. When accelerating, the force is acting at ground level, and assuming a non-zero but positive CoG height, there must be rearward weight transfer. If you throw away your internal combustion engine, and strap a jet engine to the roof, there will now be forward weight transfer under acceleration.

Likewise with deceleration, the brakes are at ground level, causing forward weight transfer. Throw that jet engine into reverse and you'll get rearward weight transfer under deceleration. The aero drag of the car is the same, it most likely acts above the CoG height, so you get more rearward weight transfer the faster you go (irrelevant of propulsion). And the rolling resistance is below the CoG height, causing the opposite.

So taking a car of 50:50 weight distribution and letting it coast down (and no engine braking) from a high speed, what will we see? The car will be rear heavy as the aero drag slows the car. Eventually the aero drag dies off and rolling resistance becomes the dominant resistance. This switch usually happens about 35mph. Assuming the aero was acting as far above the CoG as the rolling resistance is below it (unlikely I agree), the car will be back at 50:50 weight distribution. And finally, the last bit of deceleration will cause the car to get front heavy (just like prodding the brakes).

The magnitude of these affects will (in most cases) be reasonably small though (relative to actually using the brakes, or firing up the afterburner). So it's nothing you would need to worry about too much. Interesting none the less.


At this point I'm hoping everything I said is accurate or I'm going to look like an idiot.
Quote from Bob Smith :Ah cheers, have added a link to that on the download page now. Well done from refraining from mentioning APIs.

To each their own, programming methods... I gave up a while ago trying to convince people that API is the way to go in VB6. I'm currently learning Java so no need anyway
Quote from Bob Smith :It's the CoG height that matters. When accelerating, the force is acting at ground level, and assuming a non-zero but positive CoG height, there must be rearward weight transfer. If you throw away your internal combustion engine, and strap a jet engine to the roof, there will now be forward weight transfer under acceleration.

I agree in principle, but the forces acting at ground level have nothing to do with weight transfer because they act on the unsprung mass of the car rather than the chassis. The weight distribution changes during acceleration/deceleration because of d'Alembert forces acting through the centre of gravity (F=ma, basically).

Quote :Likewise with deceleration, the brakes are at ground level, causing forward weight transfer. Throw that jet engine into reverse and you'll get rearward weight transfer under deceleration. The aero drag of the car is the same, it most likely acts above the CoG height, so you get more rearward weight transfer the faster you go (irrelevant of propulsion). And the rolling resistance is below the CoG height, causing the opposite.

So you're saying that any drag forces acting below the CoG height (like the front wing) cause forward weight transfer? Surely the vertical position of the drag force relative to the pitch centre of the car is the key factor?

Forget acceleration for a moment and consider a car moving at constant speed. The drag force is not acting through either axle so the weight supported by each axle must change in response to the drag.
Including the tractive forces at the wheels (required to balance the drag since there is no acceleration) only applies if you're considering the whole car as a free body. When calculating weight transfer you only take the sprung mass of the car as a free body, therefore any forces acting on the unsprung mass do not come into consideration.

When you add acceleration, the weight transfer can now be modeled using d'Alembert forces acting on the CoG of the car where the force is equal to mass times acceleration (good old F=ma). Tractive force at the ground never comes into it, except to calculate what the acceleration of the whole car should be in response to a given applied wheel torque.

Quote :At this point I'm hoping everything I said is accurate or I'm going to look like an idiot.

Ditto! I'm really not quite sure about the physics of the situation but something seems wrong with your explanation to me.
Hmm, I was ignoring that suspension existed at all for now. Unsprung and sprung masses, roll and pitch centers, etc, are likely to complicate things.

I can see where you're coming from with what you're saying but I think I'm going refrain from commenting further until we get a third oppinion. Where are JB and Colcob hiding?
#33 - Smp
Hi Bob,
I have couple of suggestions:

1. Could you extend "0-30, 0-60, 0-150 mph" values from "N.Ns" to "N.NNs" (in order to see changes 1/100sec), please.

2. When I watch the "Acceleration Data Table" and adjust gear ratios (for ex. 1st and/or 2nd gear) it is inconvenient to (RE)scroll-down data window to track changes at higher speeds (somewhere in the bottom of those text window), cause when data changes scroller goes up automatically.
It would be good to keep scroller's position.

2.5. It Also would be convenient to make LFS GRC window resizeable.
We could use maximize window then, and to see a lot more data from "Acceleration Data Table" . Graphs would be more detail/precise too

3rd - Is a kind of bug? :
When you start LFS GRC (I used v.2.5.1), you can not access Acceleration Graph
(these one


I've found the only way to make it available is to go to "Options -> Misc -> " and to uncheck (despite it was checked by default) "Enable Acceleration calc..s" checkbox and then put a check back.
Though, you should do it every application restart

Thank you,
Smp
Quote from Smp :Hi Bob,
I have couple of suggestions:

Yay, feedback.

Quote from Smp :
1. Could you extend "0-30, 0-60, 0-150 mph" values from "N.Ns" to "N.NNs" (in order to see changes 1/100sec), please.

Hmm, well there are two issues there. Due to limitations of a text box, I can only put about 900 lines in it. So with a timestep of 0.1 seconds, that's 90 seconds of acceleration time, which is just enough to get the slow vehicles to reach top speed.
There are two solutions:
1) Allow the time step to be manually reduced, so that low speed acceleration can be displayed to more precision, but then you won't be able to test high speed acceleration as the loop will end.
2) Allow the loop to run for longer with the smaller timestep, but only output every nth line (in this case, every 10th) to the textbox... the downside here is that it will make the program even more sluggish than it already is. When are 10GHz processors coming?... The good news there is that it's mainly the outputting of text that takes time anyway, not the actual maths, so it wouldn't be ten times slower, maybe only half.

Quote from Smp :
2. When I watch the "Acceleration Data Table" and adjust gear ratios (for ex. 1st and/or 2nd gear) it is inconvenient to (RE)scroll-down data window to track changes at higher speeds (somewhere in the bottom of those text window), cause when data changes scroller goes up automatically.
It would be good to keep scroller's position.

OK I've an idea how that could be done, I'll see what I can do.

Quote from Smp :
2.5. It Also would be convenient to make LFS GRC window resizeable.
We could use maximize window then, and to see a lot more data from "Acceleration Data Table" . Graphs would be more detail/precise too

Well at the moment I can make the window resizeable but nothing will grow, you'll just have blank space. The next version is going to have to be slightly larger anyway (to fit all those silly long German words in), and with any luck you'll be able to use it in your native language too.

Quote from Smp :
3rd - Is a kind of bug? :

Weird. Works fine for me. What value is saved in the options.ini file next to "Acceleration:" (it's the very last line of the ini file)?
OK v2.5.2 is ready, multi-language support being the most notable change.

Smp: I've done something about most of your points.

I would say visit my website to see a full list of changes but my ISP is being gay and not letting me upload the devlog page for the minute (yes, they have homosexual webservers). So the full differences since the last update are:

* GRC is now translatable, ships with translations in Czech, Dutch, German and Finnish
* Acceleration times are now caclulated to 1/20th of a second
* Revised clutch simulation (slightly less simplistic)
* Altered undertray effect location (hopefully more accurate)
* The UF1000 is now available in open roof guise
* Cars now have independant idle speeds
* New GRC-style car skins for the road cars (by yours truly... apologies for my lack of creativity)
* Some minor bug fixes and improvements
Has anyone tweaked a setup sucesffully using this? I have been playing with the XFI @ Blackwood and have cured the lack of top speed, but the gearing is now in th wrong place, i.e. I'm having to upshift from 3rd to 4th on bends at 65mph rather than driving out of them in one gear.

As you can save settings & export - is there a repositry for sorted ones to save muppets like me re-inventing the wheel...?
Quote from Seahorse :Has anyone tweaked a setup sucesffully using this?

I would hope so, that's why I've bothered to continue the development for the past 18 months.

Quote from Seahorse :I have been playing with the XFI @ Blackwood and have cured the lack of top speed, but the gearing is now in th wrong place, i.e. I'm having to upshift from 3rd to 4th on bends at 65mph rather than driving out of them in one gear.

So lengthen the earlier gears without changing the 5th/final drive.

Quote from Seahorse :As you can save settings & export - is there a repositry for sorted ones to save muppets like me re-inventing the wheel...?

I have not created one, but I admit a few example settings could well prove useful. I'll add that to my to-do list and will try to include some in the next update.
Excellent, thanks Bob.

On the engine v car speed graph - Should I be aiming to have gears lines end at/above/below the upper pale green line?
Due to the way the optimum shift point is calculated, the line for any gear will always go past the green line (rpm of peak power) and drop below it when starting the next gear.

The height of the vertical line (the rpm drop) merely shows how close the gears are. (note that changing final drive doesn't affect the drop, since you aren't actually making the gears closer, but reducing them all by exactly the same percentage)


(woo, 1000 posts)
Quote from Bob Smith :
Weird. Works fine for me. What value is saved in the options.ini file next to "Acceleration:" (it's the very last line of the ini file)?

I just tried this calculator, and it is very good, but I also cannot view the acceleration graphs. And when I change some of the gears, the data in the "Acceleration" field (0-50, 0-100, 0-150 etc..) doesn't change.

In the options.ini it is written "Acceleration = True". Changing this to False with Windows notepad doesn't change anything. In the options menu in the program, if I uncheck the mark for "Enable Acceleration Calculations (Slow!)", the program crashes and I get an error message "Run-time error '383': 'Text' property is read-only".

I uploaded a screenshot plus the error message here:
http://home.online.no/~orjanbye/f1/screenshot_1.png
http://home.online.no/~orjanbye/f1/error_message.png

I have checked all dll's and ocx's etc. and even tried to replace them with the same copies found in the zip-file on the download page. I also tried to reinstall Visual basic 6.0, but nothing helped. I run Windows XP, and I tried it on another XP machine at my office, but the same result...

A possible clue to what may be wrong is that the save/load options function doesn't seem to work correctly. If I do some changes in the options menu and choose "save options", then close the program and start again, some of the option changes is gone and I have to do them again every time I use the program... Maybe I have to place the program in some default directory like C:\lfsgrc or something, otherwise the program gets confused?? I don't know, but I think this should be examined for possible bugs.


As for suggestions, I think I have one: In the Torque graph, there is an option for plotting the engine torque. But more useful would be to plot the aerodynamic drag. This is possible if you change the X-axis from wheel rpm to car speed (multiply rpm with wheel circumference?). Next is to change "torque at wheels" to "wheel thrust" by dividing torque with wheel radius. You then have a graph with Newton as unit on the Y-axis, and can draw the aerodynamic drag (also in Newton) in the same graph. Then you can see better how much force is available for acceleration as drag increase with speed. Maybe it is even possible to make an option for instead plotting the difference between the two as "net acceleration force"?

Just an idea from me, feel free to use it or scrap it.
Hi

That's basically what the tractive effort plot would show.

As for the bug, it has been reported before, however in order to fix it, I need to be able to reproduce it my end. And that's the prolem, everything works fine for me. If somebody can point to a system setting that causes the problem, perhaps I can change it my end, analyse the problem, and come up with a work around. I'm hoping it's only a smaller percentage of users that have the problem. Not many people seem to complain.

Thanks for you interest.


Edit: I suspect it might be some regional setting.
Quote from Bob Smith :Edit: I suspect it might be some regional setting.

Guess what, your spot on right!
I just tried to change the regional settings to "England, UK" and suddenly, like magic, I can see all the graphs!
Norway and England has different symbols for decimals, list separation etc., like "." instead of ","
Mystery solved!
Hopefully it is not too difficult to solve it in the software, but I know nothing about that. Except I remember having some trouble with excel-sheets many years ago for a similar reason.

Anyway, thank you for this interesting and useful tool!

(Edit): This is the setting I changed to make it work:
->Control Panel->Regional and language options:
http://home.online.no/~orjanbye/f1/regional_settings_2.png
Ah that's good to know, I'll have a play tomorrow and see if I can understand what's going on.

Cheers
Quote from MARSH2a :Will you be includig the BF1 soon?

This weekend, time pending. Adding a 7th gear is a fairly large task, there's loads of code to be tinkered with - all the torque stuff, max revs, the acceleration testing, all the graphs, the setup file format, the list goes on.
OK done some investiagtion - do "Usann" and "Sann" mean true and false in Norweigan? No idea where it's getting those words from, and only sometimes. :S Bloody MS software.
OK v2.5.3 has been updated for patches S/T/U

Atm I have not added a 7th gear for the BMW Sauber, this is just a quick update

Since my site is down atm, here is a temporary download link on my old webspace: http://myweb.tiscali.co.uk/the ... et/files/LFSGRCv2-5-3.zip

List of changes:
* Car specs updated to reflect changes in LFS Patch S
* Tyre grip values updated to reflect changed in LFS Patch S
* BMW Sauber car added
* Open top UF1000 removed - specs had been made identical to UF1000 (closed), so was redundant
* Accuracy of aero affect on car balance now improved greatly
* Tried to flatten the torque curves further for the BMW Sauber but my alogirthm still can't reproduce the torque curve for this car :-[
* Car list re-ordered
* More international support: user can now press "," (as well ".") for use as a decimal point without it being ignored
* Web address updated to point at my new URL
is this supposed to work stand alone or do i have to extract the files in the lfs folder?

when i try and run it i get an error saying:
component COMDLG32.OCX or one of its dependencies not correctly registered; a file is missing or invalid

have no idea what it means
It is stand alone - if you are missing the VB runtime files, just google the file missing and download it and put it in the same folder as GRC for now. There is a zip on my site containing all the files needed which you could grab, but my site isn't back online yet (the webspace is up but the domain doesn't work yet). Will be up at somepoint tomorrow once I've finished uploading everything again.
I just want to say that I did not know for this program before, I download it few days ago, and its wunderfull.

Great job Bob Smith, and others!!!!!!

LFS Gear Ratio Calculator 2½ [deceased]
(76 posts, started )
FGED GREDG RDFGDR GSFDG