|
|
12-11-2007, 11:41 PM | #221 | ||
Colonel
99
Rep 2,000
Posts |
Quote:
When I (rather politely, I thought) mentioned that the numbers were well off, but the only thing I could put my finger on was something at the top end, you did your normal ego protection attack, including this gem: "Why are these numbers “reasonable”? Simply because they are typically within the average numbers reported from tests plus or minus the variation in the reported test numbers. This is the hallmark of an acceptable/reasonable simulation." Your bolding, by the way. Down here at the track, though, four tenths is four car lengths, and you just got your ass kicked. Quote:
Bruce |
||
Appreciate
0
|
12-12-2007, 03:49 AM | #222 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
Any monkey can tune a simulation to match results. You can tune the data entries for an Enzo to match performance numbers for a Hugo. Ever heard of "garbage in garbage out" a classic expression in the world of simulation. There is no skill nor magic in that. Other than weight, shift times, transmission losses and some basic launch optimization, the numbers I used for the input E46 M3 simulations were blind/default and completely un-tuned. Do you know what that means? NO OTHER DATA ABOUT THAT PARTICULAR CAR NOR TEST THAT YEILDED THAT PARTICULAR SUBSET OF RESULTS WAS KNOWN. In these circumstances the results I obtained are again within the accuracy of the variability of known tests. I just can not say that enough times. I am seriously tiring of trying to help you understand. I actually think this will be my last exchange with you on this topic. I've simply run out of energy, explanations, analogies, logic, patience, etc. You have missed the boat. If I recall correctly gbb357 did not ask me to engage in a curve fitting exercise i.e. to "model" a single set of test results. That is about as contradictory and nonsensical as you can get - to "model" a single set of results. You do not model one test case unless you are modeling a process that is highly repeatable. I am honestly begining to loose all hope of getting the basics concepts of simulation, accuracy, random and systematic errors and predictability across to you. If gbb357 did in fact want to test my curve fitting abilities (rather than my ability or the quality of CarTest) I can whip out a CarTest simulation pretty quickly to match those exact test results. What then happens when friendly forum user X presents me with his test results or the results from magazine Y that are drastically different????? Well I could go tweak the simulation to match those results as well but you simply must see the logical consequence of this repeated ad infinitum. The real problem with that process would simply be the lack of justification for changing the parameters to tune to the results each time. In simulation you should either have a justification for your parameters or bracket/envelope them as I demonstrated quite effectively in my use of simulation to prove the C63 AMG is under-rated. You did see this exchange in the "Top Gear..." thread, right? Did you notice how incredibly close one data point on how over-rated the CLK63 AMG is corresponded to my power prediction of the C63 AMG required to get the performance it does (note almost identical engine)? This is a much better use of simulation than trying to obtain a very highly accurate absolute measurement of any particular single metric or groups of metrics for a single car. Not sure how many times I have to repeat this as well but simulation (esp. as I am using it, to try to match a "moving target" of varying magazine results) is actually much more accurate in relative predictions rather than absolute predictions. Specifically, the time difference between two cars in the 0-100 or trap difference in the quarter, etc. This is because a relative simulation comparison removes some of the systematic errors present in both results. Almost magic! Even with this ranking of relative over absolute, and of bracketing/enveloping over absolute, I am still happy with the absolute results I have obtained and they are reasonable compared to the spread of test results. Indeed the accuracy I have obtained thus far in my efforts are not suitable to predict with absolute accuracy results for a drag racing team. Like I explained many posts ago that is a fairly different goal, process and simulation strategy. You simply never read nor understand (nor rarely reply directly to) what I post. All of this being said I do think that 1/4 mi trap speed may be the worst results for my absolute predictions from my use of CarTest. Like enigma mentioned this appears to be related to some air drag effects or other higher speed dominated effects. Bruce, neither CarTest nor my simulations are perfect, but at least I understand the limitations of both and the entire concept of predictability. You have a lot of homework and learning to do here. Get busy dusting off the TRS-80. I am truly done with you on this topic. Adios. |
|
Appreciate
0
|
12-12-2007, 04:18 AM | #223 |
Major General
1207
Rep 8,034
Posts |
I was only highlighting that is the majority of the reviews between the E46 M3 and the RS4 the Audi came out on top and in quite a few of reviews between the RS4 and the E92 M3 the same has being true. What you or I think is irrelevant, my comments where directed at roadtesters opinions not ours.
|
Appreciate
0
|
12-12-2007, 04:40 AM | #224 | |
Major General
1207
Rep 8,034
Posts |
Quote:
But I do agree that the RS4 is much easier to drive quickly and I have said as much on numerous occasions, the same applies to all AWD cars and not just Audis. |
|
Appreciate
0
|
12-12-2007, 04:12 PM | #225 | ||
Captain
72
Rep 706
Posts |
I thought this was funny too Bruce.
Quote:
|
||
Appreciate
0
|
12-12-2007, 04:35 PM | #226 | |
Major General
1207
Rep 8,034
Posts |
Quote:
I don't think that will go down to well, best stay out of the firing line. As for the times listed above for the IS-F, in Europe I believe the average for the M3 is something like 4.7s to 100km/h so if you re-adjust back to 60mph that will probably end up at 4.6s, based on that the two cars are all but equal though this just up to speeds of 60mph, the Lexus should pull harder after this speed what with it's eight speed gearbox. Has anyone an opinion on what the gap would be between the two at say 150mph ............. (no simulators required). |
|
Appreciate
0
|
12-12-2007, 05:01 PM | #227 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
It is simply ludicrous to believe that one update of one input parameter from a known incorrect value to a known correct one is not the best course of action. toothdoc was nice enough to point out the error, I fixed it and the results are now nearly perfect considering the average of the 0-60 figures as you have reported them vs. the blind simulation with std. transmission loss parameters. I can not believe how badly and desperately you guys are struggling to find something to nitpick at here. The results are what they are and they continue to validation the tool and my applicaiton of it. Have fun - your petty nitpicking seems childish, boring and entirely non-productive to me. |
|
Appreciate
0
|
12-12-2007, 05:17 PM | #228 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
Very curious as to how others come up with their "guesses" or "opinions". |
|
Appreciate
0
|
12-12-2007, 05:27 PM | #229 | ||
Major General
1207
Rep 8,034
Posts |
Quote:
What I wanted to know was between the IS-F and the current M3 in manual form. Quote:
|
||
Appreciate
0
|
12-12-2007, 05:44 PM | #230 |
Lieutenant General
634
Rep 10,407
Posts |
|
Appreciate
0
|
12-12-2007, 05:52 PM | #231 |
Major General
1207
Rep 8,034
Posts |
That is indeed quite close. In fact when you calculate how much a car would have travelled in one second at this speed, four lengths of either of these cars would mean the difference between the two times would only be 0.2~0.3s.
That is even closer than I would have thought. Unless my calculation are way off, of so can someone explain what I am doing wrong. |
Appreciate
0
|
12-12-2007, 06:47 PM | #232 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
|
|
Appreciate
0
|
12-12-2007, 11:38 PM | #233 |
Colonel
99
Rep 2,000
Posts |
An update on the IS-F quarter mile
Since I've been one of the principal hijackers of this thread (sorry, folks), I thought perhaps I could take the opportunity to post something on topic.
Figuring what the heck, I have "Quarter, Jr.", and since it seemed to do so well when tracking the E46 M3, I thought I'd hurl in the IS-F data from the current Car & Driver test and see what it hurled back out at me. Here's the result: 60' 2.06 330' 5.53 660' 8.33 MPH 88.6 1000' 10.71 1/4 12.72 MPH 112.3 Pretty good stuff. C & D got 12.7 @ 114, so even though my calculated MPH is a little low, that discrepancy represents only about a 2% power differential (about eight horsepower), since C & D calculates finsh line speed as far as I know, which Quarter, Jr. shows at 113.2. In addition, the model shows an 0-60 time of 4.39 seconds, and an 0-100 time of 10 seconds flat, each of which is two tenths of a second off the C & D numbers. I used a car weight of 3970 pounds assuming driver and their hot-damn test gear, and again used the SAE (gross) Standard Day weather conditions, which is approximately what they use. Combined with the dyno results quoted long ago and far away in this string, I'd say a case could be made that although the Lexus is a healthy runner, it is probably not significantly under rated in terms of power, contrary to previous testimony by He Who Must Not Be Named. OK, that's not fair, since Swamp has sworn to be mute. Fun, though. Just as an additional check, I entered the BMW 328 data from the same issue, and got a 14.83 @ 95.4, against their 14.8 at 95. Hm-m. Now I'm a bit high on MPH, with a finsh line speed of 96.1. That represents about a 3.5% increase in calculated power vs what C & D got, which is also around eight horsepower. Results seem very close to a more or less known quantity like C & D, but not necessarily against other publications who's test methods are unknown to me. Edit: Note that there isn't any real tuning to be done with this very simple product. When I get ShiftMaster working, I'll try a couple of passes as well. Bruce |
Appreciate
0
|
12-13-2007, 03:13 AM | #234 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
I am curious about a few key parameters that you used in your sim, namely: -Automatic transmission losses -Torque converter losses, if any -Shift times -Did you input a torque or hp curve or just peaks Also why are you quoting two simulation results for trap speeds? Lastly do you really have any reason to believe your sims should match C&D better than any other magazines results or is it simply convenient? |
|
Appreciate
0
|
12-13-2007, 05:33 AM | #235 |
Major General
1207
Rep 8,034
Posts |
I am very curious as to how exact these can be, like do they allow the input of surface types, awd transmissions etc. or is this too much to expect.
P.S. Autocar have driven the IS-F and it's in this week's magazine, their belief time with the left them feeling it's not the equal dynamically of the RS4, M3 or C63 but like the engine and it's performance is very impressive..........not so the interior or the UK's suggested retail price. Also in same issue they did a proper roadtest on the C63 on UK roads and rate it now No1 ahead of the M3. When I get to read in full will give full run down of points, that's if someone hasn't beaten me to it. |
Appreciate
0
|
12-13-2007, 10:16 AM | #236 | ||
Colonel
99
Rep 2,000
Posts |
Quote:
This is either infuriating or amazing, depending on your point of view. I would say in addition that Patrick Hale of RSA was one of those resident geniuses at the time we were messing about with this stuff, and "Quarter" was pretty much a legendary package. If memory serves, Quarter demands a good deal more input than Quarter, Jr., but I'd guess that the Jr. package still uses much of that magical original code, with obvious defaults for necessary parameters. I'm quoting trap speed and speed at the finish line, and these have been somewhat different since around 1990, I think. That's when the NHRA got nervous about Top Fuelers getting to be too damned fast, and thus ending up in the boondocks before they could get the things stopped on tracks with shorter shutdown areas. They changed the speed trap from a 132 foot block bracketing the finish line to a 66 foot block ending at the finish line. At the time, I remember analyzing over 100 published road tests before coming up with the idea that the difference averages about seven tenths of one percent (.007) for most cars. Anal? Who me? In any event, Quarter, Jr. gives you both speeds. Quote:
Bruce PS - While we're sort of locked into the "She don't idle too good but she's hell in the quarter" mindset, everybody should get themselves to a drag strip at least once in their lives to watch the Top Fuelers run. I hadn't been to a strip in perhaps thirty years or so to actually watch an event (just attending for time trials for whatever I was driving at the time), but my bride and I went to watch a national NHRA event a couple of years ago. My theory is that you don't actually watch the Top Fuelers run. They happen to you. |
||
Appreciate
0
|
12-13-2007, 12:29 PM | #237 | |
Lieutenant General
634
Rep 10,407
Posts |
Quote:
+1 on watching top fuelers, esp. the rails. Even just the acoustics is enough to raise every hair on your neck and give you goosebumps. The pits are quite a hoot as well with the nuts wearing gas masks and industrial ear protection getting within feel of flame throwing headers belching noxious yellow nitro-methane. |
|
Appreciate
0
|
12-13-2007, 06:08 PM | #238 | |||
Colonel
99
Rep 2,000
Posts |
Footie, sorry I'm so late with this reply.
Quote:
As a for instance, I personally believe that on an E46 M3, a change in final drive ratio to a 4.10 from a 3.62 would be a good one. In everyday driving out on the highways, it would be a more responsive and enjoyable beast, answering your right foot with more eagerness. Cruise rpm would jump up from, say, 3000 to about 3400 (at around 80 mph), so the car would be making that much more power. The change would come at the cost of some extra fuel consumption and possibly noise, although these engines are very smooth at cruise, so noise won't be much of an issue. Quote:
That said, are we talking about M5 vs M6? I thought that they had identical engines and drivetrains, with the M5 having a minor weight penalty. Obviously I've got it wrong, so let me know what we actually *are* comparing. Quote:
In any event, I used the example of a waterwheel I observed way back when, making around 2600 foot pounds of torque according to the folks in the old-timey flour mill (and tourist attraction) that the wheel served. Now, there's no way your M3, or C63, S4 or S5 is going to be making anywhere near that amount of torque, but the rest of the story is that the wheel was rotating at about 12 rpm. If you hooked that wheel to the drive wheels of a car, that car would lurch up to 12 rpm of the drive wheels in an instant, and the waterwheel would hardly notice. However, 12 rpm of the drive wheels is about one mile per hour for the average car, and what if you wanted to go faster? Well, you'd have to gear it up. If you wanted to go 60 miles per hour, you'd need to gear it up by 60 times, and returning to the topic of simple machines you studied as a kid, you'd have one 1/60th of the torque you had without gearing, meaning you'd have about 43 foot pounds of torque to keep the car at 60. Ain't enough, as we English majors say, and if you do your HP/Torque sums, 12 times 2600, over 5252 gives you......6 HP. Horsepower is essentially torque at speed, and as speed increases, so does the need for power, in a completely linear way. Remember how that water wheel booted you up to one mph? Well, if you're addicted to our Lord God of Acceleration, you're going to need 360 HP if you want to be booted just as hard at 60 - and that's not even including increased air and rolling resistance. What causes some confusion is that, in any given gear, the car will accelerate in a way that exactly matches it's torque curve, so it's not accelerating as hard at the power peak as at the torque peak. At 7900 rpm, the M3 is down to 221 foot pounds of torque, so it's accelerating only 84% as hard as it did at the torque peak in that gear. However, even though torque is what is actually accelerating the car, making more power at any given car speed is the key. As an example, an E46 M3 is doing something around 40 miles per hour in second gear at its 4900 rpm torque peak, so in order to get a feel for acceleration potential at that speed in second gear, you'd multiply 263 foot pounds of torque by the second gear ratio of 2.53, times the final drive ratio of 3.62, which gives you 2408 foot pounds of torque at the drive wheels - enough to sock you in the back pretty hard. In fact, as mentioned, that's as hard as you're going to accelerate in second gear. At that point, by the way, you're making 245 horsepower. However, what if the car was making max power at that point instead of max torque? In order to do that, you'd need to re-gear the car in order to get to its 7900 rpm power peak at 40 in second gear, so you'd need a 5.84 final drive ratio. There, you'd have about 221 foot pounds of torque (in the U.S., at 333 HP) times the second gear ratio of 2.53, times 5.84, giving you TA DA! 3265 foot pounds at the drive wheels - or about a 36% harder boot in the back. Just doubling back here, 333 horsepower is about 36% more than 245 horsepower. What're the odds? Torque is what you feel, but horsepower rules. My dad once told me (quite truthfully, I think) that the way you could tell who was a true expert in something is that he or she could relate fairly complex ideas in simple terms, using relatively few words. By that standard, I'm more or less an idiot - but, did all that rambling actually add anything to your knowledge? Bruce PS - All this said, if you have two cars with pretty even power to weight ratios, the car with better torque to weight will win, because it'll be making more power to weight on its way up to the power range off the line, and also after each shift. |
|||
Appreciate
0
|
12-13-2007, 06:20 PM | #239 | |
Captain
13
Rep 689
Posts |
Quote:
On a road course gear changes can be a win by keeping the car in the ideal power band and avoiding having to shift at all the wrong places. I swear in my car I ht the rev limiter about 300' before 3 diffrent brake zones at Laguna. Only in issue in a MT car, in the DCT you would go ahead and do the shifts. |
|
Appreciate
0
|
12-14-2007, 04:22 AM | #240 |
Major General
1207
Rep 8,034
Posts |
Bruce,
I roughly get it but what happens when you come across a car like the new RS6, it's peak power of 580HP @ 6250~6750rpm and torque of 650Nm @ 1750~6250rpm (a peak which is higher than even the M5's 6100rpm). Surely figures like these it has to be accelerating hard than the M5 or is it a case that at this point the gearbox choices come into play. It just I have read on numerous sites that the debate seem to be split pretty evenly on which will be the quickest. I too would love an opinion from someone with knowledgeable experience on this, because my heart says (being a power hungry mother) that surely the RS6 will be way quicker what with it's extra power and torque but then my head (capt sensible) chimes in saying things like weight, 7 gears and higher peak power etc. Some Bruce, what do you think? |
Appreciate
0
|
12-14-2007, 08:21 AM | #241 | |
Colonel
99
Rep 2,000
Posts |
Quote:
The BMW has a more efficient and "better" gearbox (I assume it will shift more quickly than the Audi), but assuming Audi will allow pre-loading off the line (and they will), the RS6 will launch like a ball bearing out of a slingshot, leaving the M5 for dead. At speed, it'll be closer, I think. I'd still give the Audi the nod, though. Way up there into triple digits, power over net frontal area gets more important than power to weight, and the Audi will get the nod there as well. Bruce PS - I have no axe to grind, here. I don't give a damn about either of the companies, nor any other auto manufacturer, for that matter. On second thought, that may not be correct. I'm actually pretty much a badge whore for all of them. Right now, for instance, I'm pretty excited about what Hyundai is doing, and watching everybody else sweat. Who said promiscuity is wrong? |
|
Appreciate
0
|
12-14-2007, 10:30 AM | #242 | |
Captain
72
Rep 706
Posts |
I'm with you on this one Bruce.
Quote:
|
|
Appreciate
0
|
Post Reply |
Bookmarks |
|
|