Dyno Dynamics - Shoot out modes
Discussion
Mignon said:
spitfire4v8 said:
Std car quoted at 301hp
dyno dynamics say 302 for a tested std car.
Mignon suggests 275hp is more realistic
Only on the basis of your tests of how much S6F seems to read above steady state. But who knows how that other dyno reads. I've seen Dasteks read differently and one that moved premises never read the same again after the move.dyno dynamics say 302 for a tested std car.
Mignon suggests 275hp is more realistic
Looking at Rototest figures BMW seem to accurate or even conservative in their power claims. They haven't tested the 135i but similar cars are certainly making the claimed numbers.
Of course we have no actual proof that the steady state mode reads true either on any particular dyno. I suppose it's not beyond the realm of possibility that it reads low on some of them and the shootout mode in some conditions corrects for that.
It was all much easier in'th olden days. When I were a lad people only had Clayton water brake rollers that only gave wheel figures and as far as I could tell from customers who tried several of them with the same car they all read near as dammit the same. A bloke from Clayton used to go round calibrating them with his company car Sierra estate I think it was. Well he used the weights and stuff first but he always gave it a run on the Sierra just to make sure he hadn't cocked things up.
I put my Mk1 Astra Gte on a bunch of dynos back in the 80s/early 90s and every one said 90 at the wheels. Dave Walker used it for a cam test article in CCC magazine once. Very expensive Shrick cam I think it was and it made not an effing scrap of difference.
Then people started wanting flywheel numbers and some idiot invented the coast down procedure and more idiots tinkered with computers massaging the numbers and it all went to hell in a handbasket. Talking of hell, if there's a highway to hell but only a staircase to heaven I guess that gives you some idea of the relative traffic numbers they were expecting.
It was all much easier in'th olden days. When I were a lad people only had Clayton water brake rollers that only gave wheel figures and as far as I could tell from customers who tried several of them with the same car they all read near as dammit the same. A bloke from Clayton used to go round calibrating them with his company car Sierra estate I think it was. Well he used the weights and stuff first but he always gave it a run on the Sierra just to make sure he hadn't cocked things up.
I put my Mk1 Astra Gte on a bunch of dynos back in the 80s/early 90s and every one said 90 at the wheels. Dave Walker used it for a cam test article in CCC magazine once. Very expensive Shrick cam I think it was and it made not an effing scrap of difference.
Then people started wanting flywheel numbers and some idiot invented the coast down procedure and more idiots tinkered with computers massaging the numbers and it all went to hell in a handbasket. Talking of hell, if there's a highway to hell but only a staircase to heaven I guess that gives you some idea of the relative traffic numbers they were expecting.
Those were the days Dave, I remember testing your Astra, with hindsight do you thinking altering the cr to suit the different cams would have made a difference?
Lee Higgins used to service and calibrate the Clayton dynos, he also serviced the Camic co/hc meters. Sadly he died late 90s. Robin from Clayton used to do the work after that but it was £75 an hour. We were flooded in 2001 and the insurance paid out for a full rebuild carried out by Robin. When Clayton stopped working on the dynos I inherited the calibration weights and the gauge for calibration. We still have them to match the Clayton engine dyno we have never used, complete with flashy digital horsepower torque and rpm readout, not a patch on the old quivering needle on the wall.
Lee Higgins used to service and calibrate the Clayton dynos, he also serviced the Camic co/hc meters. Sadly he died late 90s. Robin from Clayton used to do the work after that but it was £75 an hour. We were flooded in 2001 and the insurance paid out for a full rebuild carried out by Robin. When Clayton stopped working on the dynos I inherited the calibration weights and the gauge for calibration. We still have them to match the Clayton engine dyno we have never used, complete with flashy digital horsepower torque and rpm readout, not a patch on the old quivering needle on the wall.
GregK2 said:
I have a log I took myself if that helps?
https://datazap.me/u/gregk2/stage2?log=0&data=...
Are you able to export that data to say a CSV file and does it also have an each piece time stamped?https://datazap.me/u/gregk2/stage2?log=0&data=...
Stan
Stan Weiss said:
GregK2 said:
I have a log I took myself if that helps?
https://datazap.me/u/gregk2/stage2?log=0&data=...
Are you able to export that data to say a CSV file and does it also have an each piece time stamped?https://datazap.me/u/gregk2/stage2?log=0&data=...
Stan
It gives time in seconds / ms. Selecting "Time" in the graph lets you see it also.
https://docs.google.com/spreadsheets/d/1dJdm-eH_hk...
Edited by GregK2 on Thursday 15th November 19:24
Stan Weiss said:
Are you able to export that data to say a CSV file and does it also have an each piece time stamped?
Stan
Look more closely at the page Stan. The data is all there in a chart. There is even a column for torque at the clutch and I have no idea how the car knows that but the numbers are not far off the dyno chart ones.Stan
GregK2 said:
Yes, it starts as a CSV that I upload to datazap to create the graph. But the CSV can be opened too.
It gives time in seconds / ms. Selecting "Time" in the graph lets you see it also.
https://docs.google.com/spreadsheets/d/1dJdm-eH_hk...
Thanks, i will have look and see if I can calculate anything from those numbers.It gives time in seconds / ms. Selecting "Time" in the graph lets you see it also.
https://docs.google.com/spreadsheets/d/1dJdm-eH_hk...
Edited by GregK2 on Thursday 15th November 19:24
Stan
If the OP wishes to find an absolutely flat bit of road and datalog a full throttle run in 3rd gear all the way through the rev range then Stan and I will tell you exactly how many horsies you have. We'll need tyre size, gear ratios and vehicle weight as tested with fuel and driver. Aero drag would be nice too but it's not hugely critical. cD and frontal area if you can find it.
A little history. 10 or more years ago I wrote a routine to take OBDII data logging and to try to produce a HP and Torque curve. The generated curves were useless without massive smoothing. So I gave up. Later I was then able to get data from an engine inertia dyno which worked very well with my code without any smoothing.
This data produces the same useless curves without massive smoothing. Here are graphs of the raw and smoothed data. I still need to spend some more time playing with a few of the programs inputs.
Stan
This data produces the same useless curves without massive smoothing. Here are graphs of the raw and smoothed data. I still need to spend some more time playing with a few of the programs inputs.
Stan
Raw data calculations
Rear Aero Rolling Rear W Accele Time Rate
RPM MPH Velocity Wheel dynamic Resist. Elapsed Horse ration Differ Rev Per
ft/sec Torque Drag - HP HP Time Power in G's ential Sec
2434.0 173.787 254.888 0.00 .000 .000 .0000 0.00 0.0000 0.0000 0.0
2446.0 174.644 256.145 67.67 .000 .000 .1010 31.52 .3867 0.1010 118.8
2458.0 175.501 257.401 62.14 .000 .000 .2110 29.08 .3551 0.1100 109.1
2474.0 176.643 259.077 81.37 .000 .000 .3230 38.33 .4650 0.1120 142.9
2484.0 177.357 260.124 52.26 .000 .000 .4320 24.72 .2986 0.1090 91.7
2500.0 178.500 261.799 84.38 .000 .000 .5400 40.17 .4822 0.1080 148.1
2514.0 179.499 263.265 71.20 .000 .000 .6520 34.08 .4068 0.1120 125.0
2525.0 180.285 264.417 58.01 .000 .000 .7600 27.89 .3315 0.1080 101.9
2538.0 181.213 265.779 67.31 .000 .000 .8700 32.53 .3847 0.1100 118.2
2554.0 182.355 267.454 89.35 .000 .000 .9720 43.45 .5106 0.1020 156.9
2568.0 183.355 268.920 73.84 .000 .000 1.0800 36.10 .4219 0.1080 129.6
2586.0 184.640 270.805 92.37 .000 .000 1.1910 45.48 .5278 0.1110 162.2
2602.0 185.782 272.481 82.85 .000 .000 1.3010 41.05 .4734 0.1100 145.5
2636.0 188.210 276.041 195.62 .000 .000 1.4000 98.18 1.1178 0.0990 343.4
2654.0 189.495 277.926 56.96 .000 .000 1.5800 28.78 .3255 0.1800 100.0
2710.0 193.494 283.791 289.97 .000 .000 1.6900 149.62 1.6570 0.1100 509.1
2747.0 196.135 287.665 210.75 .000 .000 1.7900 110.23 1.2043 0.1000 370.0
2806.0 200.348 293.844 300.05 .000 .000 1.9020 160.31 1.7146 0.1120 526.8
2855.0 203.847 298.975 284.79 .000 .000 2.0000 154.81 1.6274 0.0980 500.0
2881.0 205.703 301.698 135.87 .000 .000 2.1090 74.53 .7764 0.1090 238.5
2979.0 212.700 311.960 502.88 .000 .000 2.2200 285.24 2.8736 0.1110 882.9
3055.0 218.126 319.919 393.53 .000 .000 2.3300 228.91 2.2488 0.1100 690.9
3106.0 221.768 325.260 264.08 .000 .000 2.4400 156.18 1.5090 0.1100 463.6
3164.0 225.909 331.333 297.62 .000 .000 2.5510 179.30 1.7007 0.1110 522.5
3229.0 230.550 338.140 373.97 .000 .000 2.6500 229.92 2.1370 0.0990 656.6
3312.0 236.476 346.832 472.76 .000 .000 2.7500 298.13 2.7015 0.1000 830.0
3382.0 241.474 354.162 394.76 .000 .000 2.8510 254.21 2.2558 0.1010 693.1
3423.0 244.402 358.456 259.48 .000 .000 2.9410 169.12 1.4827 0.0900 455.6
3500.0 249.899 366.519 438.58 .000 .000 3.0410 292.28 2.5062 0.1000 770.0
3578.0 255.469 374.687 407.60 .000 .000 3.1500 277.68 2.3291 0.1090 715.6
3648.0 260.467 382.018 359.20 .000 .000 3.2610 249.50 2.0526 0.1110 630.6
3730.0 266.321 390.605 467.06 .000 .000 3.3610 331.71 2.6689 0.1000 820.0
3769.0 269.106 394.689 201.94 .000 .000 3.4710 144.92 1.1540 0.1100 354.5
3854.0 275.175 403.590 474.66 .000 .000 3.5730 348.31 2.7123 0.1020 833.3
3929.0 280.530 411.444 435.91 .000 .000 3.6710 326.10 2.4909 0.0980 765.3
3967.0 283.243 415.423 216.44 .000 .000 3.7710 163.49 1.2368 0.1000 380.0
4031.0 287.813 422.125 328.41 .000 .000 3.8820 252.06 1.8766 0.1110 576.6
4114.0 293.739 430.817 433.72 .000 .000 3.9910 339.74 2.4784 0.1090 761.5
4169.0 297.666 436.577 313.27 .000 .000 4.0910 248.67 1.7901 0.1000 550.0
4246.0 303.164 444.640 434.24 .000 .000 4.1920 351.06 2.4814 0.1010 762.4
4302.0 307.162 450.504 318.97 .000 .000 4.2920 261.27 1.8227 0.1000 560.0
4354.0 310.875 455.950 299.18 .000 .000 4.3910 248.02 1.7096 0.0990 525.3
4420.0 315.587 462.861 372.21 .000 .000 4.4920 313.24 2.1269 0.1010 653.5
4480.0 319.871 469.145 345.20 .000 .000 4.5910 294.46 1.9726 0.0990 606.1
4554.0 325.155 476.894 383.18 .000 .000 4.7010 332.25 2.1896 0.1100 672.7
4604.0 328.725 482.130 256.57 .000 .000 4.8120 224.92 1.4661 0.1110 450.5
4672.0 333.580 489.251 355.34 .000 .000 4.9210 316.10 2.0305 0.1090 623.9
4740.0 338.435 496.372 348.94 .000 .000 5.0320 314.92 1.9939 0.1110 612.6
4806.0 343.148 503.283 341.75 .000 .000 5.1420 312.73 1.9529 0.1100 600.0
4861.0 347.075 509.043 310.17 .000 .000 5.2430 287.08 1.7724 0.1010 544.6
4912.0 350.716 514.383 268.97 .000 .000 5.3510 251.56 1.5370 0.1080 472.2
4974.0 355.143 520.876 353.14 .000 .000 5.4510 334.45 2.0180 0.1000 620.0
5042.0 359.998 527.997 348.94 .000 .000 5.5620 334.98 1.9939 0.1110 612.6
5110.0 364.853 535.118 387.32 .000 .000 5.6620 376.85 2.2133 0.1000 680.0
5134.0 366.567 537.631 135.35 .000 .000 5.7630 132.31 .7734 0.1010 237.6
5194.0 370.851 543.914 310.68 .000 .000 5.8730 307.25 1.7753 0.1100 545.5
5262.0 375.706 551.035 358.63 .000 .000 5.9810 359.31 2.0493 0.1080 629.6
5323.0 380.061 557.423 313.02 .000 .000 6.0920 317.25 1.7887 0.1110 549.5
5376.0 383.846 562.973 298.89 .000 .000 6.1930 305.95 1.7080 0.1010 524.8
5415.0 386.630 567.057 226.67 .000 .000 6.2910 233.71 1.2953 0.0980 398.0
5470.0 390.557 572.817 279.71 .000 .000 6.4030 291.32 1.5983 0.1120 491.1
5534.0 395.127 579.519 331.40 .000 .000 6.5130 349.19 1.8937 0.1100 581.8
5570.0 397.697 583.289 209.24 .000 .000 6.6110 221.91 1.1956 0.0980 367.3
5610.0 400.553 587.478 227.84 .000 .000 6.7110 243.37 1.3019 0.1000 400.0
5690.0 406.265 595.855 414.25 .000 .000 6.8210 448.79 2.3671 0.1100 727.3
5710.0 407.693 597.950 101.71 .000 .000 6.9330 110.58 .5812 0.1120 178.6
5780.0 412.691 605.280 402.74 .000 .000 7.0320 443.23 2.3014 0.0990 707.1
5849.0 417.618 612.506 396.99 .000 .000 7.1310 442.11 2.2685 0.0990 697.0
5854.0 417.975 613.029 25.66 .000 .000 7.2420 28.60 .1466 0.1110 45.0
5912.0 422.116 619.103 297.62 .000 .000 7.3530 335.02 1.7007 0.1110 522.5
5982.0 427.114 626.434 362.47 .000 .000 7.4630 412.85 2.0712 0.1100 636.4
6014.0 429.399 629.785 154.46 .000 .000 7.5810 176.88 .8827 0.1180 271.2
6058.0 432.540 634.392 229.93 .000 .000 7.6900 265.21 1.3139 0.1090 403.7
6099.0 435.468 638.686 210.39 .000 .000 7.8010 244.32 1.2022 0.1110 369.4
Averages 271.29 221.91 0.1069 476.3
Smoothed
Torque
67.674
67.674
62.137
69.478
68.665
69.757
70.115
70.072
73.635
77.228
87.152
95.857
114.766
140.697
166.513
207.624
226.753
256.146
284.855
298.030
330.946
337.548
343.258
358.430
361.108
375.882
381.631
376.990
387.706
379.595
379.748
386.917
365.450
366.676
361.152
343.170
354.463
357.262
354.210
362.507
351.626
345.349
344.817
339.549
340.078
334.005
333.945
329.541
325.167
323.831
323.967
321.149
315.415
309.837
294.794
298.323
299.256
294.650
295.837
281.199
271.885
275.911
264.433
270.653
287.372
271.396
282.020
278.032
248.315
255.651
245.291
154.465
229.925
210.388
.000
Edited by Stan Weiss on Thursday 15th November 22:07
Edited by Stan Weiss on Thursday 15th November 22:08
spitfire4v8 said:
As an aside .. are there any motor manufacturers or engine tuners/builders who actually do publish accurate information ? You'd natually assume that BMW might publish accurate info, but I don't know.
I'm not involved myself, but I understand there are regulations about how the published figure is measured, including things like correcting to established environmental conditions and various tolerances for the accuracy of the measurement equipment. I've been told the manufacturers game the system quite extensively and will measure with equipment and under conditions which allow them to bias the results in their favour. Mignon said:
Most of it isn't at full throttle Stan so there's not much to deduce from it.
My bad. I missed that the pedal position was 100% and the throttle is drive by wire so that only opens as much as needed. Assuming that was a flat road I obtained a perfect acceleration curve matching actual to calculated 10 mph increments to within 0.01 seconds for every increment. I've never actually obtained a curve as perfect as that right the way through a gear.At the flywheel you have peak torque of 417 ft lbs at about 3000 to 3500 rpm and peak power is 357 at about 5000 to 5250. That's 303 bhp at the wheels.
From the intergoogles I obtained curb weight of 3420 lbs so plus driver I used 3600. Final drive 3.08, 3rd gear 1.58, rear tyres 245/35/18. MPH per 1000 rpm in 3rd is 14.53. Frontal area 22.4 sq ft, cD 0.33. Spec sheet is here.
http://media.caranddriver.com/files/2008-bmw-135i-...
The car is very traction limited, it'll light the tyres up without even trying so acceleration off the line depends on tyres and launch but on very good road tyres I get 0-60 mph in 4.7s. 1/4 mile is 13.26 @ 110 mph. Top speed is 171.
A couple more minor points. My computer system calculated the boost pressure it would expect from the torque curve it calculated to match the actual acceleration. At peak torque at 3000 rpm it calculated 13.8 psi. Your datalog says exactly 13.8 psi. It's actually overshooting the target boost there by quite a bit and the timed acceleration shows that up very clearly. At peak power at 5000 rpm the computer calculates 12.8 psi and the datalog says 12.9 psi. So I can be very confident that the computer's calculated power curve is exactly matching not only the timed acceleration but also the actual boost pressures generated. Above 3500 rpm the actual boost is lagging the target boost a bit. The final calculation the computer did was the power curve if you could achieve target boost and it says 367 to 370 bhp at 5100 rpm.
With the cross check of not only timed performance but boost pressure curves and the 0.01 second match between calculated and actual times I can rate that simulation to 1% accuracy. Less than 4 bhp or 4 ft lbs error assuming the car weight I've used is correct and there was no gradient involved.
Using actual performance and Newtonian physics is actually far more accurate than rolling roads or even engine dynos which as we've seen have too much buggering about with the numbers going on inside them. F=ma doesn't have any error built in unless you're travelling at relativistic speeds or in a very strong gravity well and I think I can discount both of those for a BMW.
With the cross check of not only timed performance but boost pressure curves and the 0.01 second match between calculated and actual times I can rate that simulation to 1% accuracy. Less than 4 bhp or 4 ft lbs error assuming the car weight I've used is correct and there was no gradient involved.
Using actual performance and Newtonian physics is actually far more accurate than rolling roads or even engine dynos which as we've seen have too much buggering about with the numbers going on inside them. F=ma doesn't have any error built in unless you're travelling at relativistic speeds or in a very strong gravity well and I think I can discount both of those for a BMW.
Fascinating stuff but it brings one back to the fact the dyno whether engine or chassis is by far a quicker way of sorting the car tuning than going up and down a theoretical empty road where there are no speed limits or time restrictions on measurement. You are still making many assumptions to make the numbers fit eg car completely warmed up, tyre pressure, width and grip and compound, road surface friction and temperatures etc etc. certainly your method of chomping numbers is cheaper than buying a dyno. but hey, I can change the timing or fuelling in a couple of minutes then get results in seconds. Using a rolling road for tuning,taking the argument of figures to the lowest level, assuming the numbers are repeatable, in statistical analysis perhaps the most powerful test is the sign test, + or - . All one does is compare pluses and minuses from the test results, make a decision and move on. Result, happy customers whether road or race. The dyno is only a tool as many of us keep saying, use that tool correctly and it is a good tool which maybe explains why most tuners use dyno testing in some guise rather than drive up and down a road. I suppose it would be fun for a dedicated person to do tests your way before and after testing to see the modelled differences.
So I cranked up my simulator today...
Same atmos. conditions as the OP
This is the car producing 330 hp @ 5500 rpm in Shoot_2R, ramping at 100
280 WHP, 330 Fly.
I then changed to Shoot_6F, without touching the simulator :
282 WHP, 339 Fly.
The 2 hp higher WHP figure (<1%) is due to the assumed higher driveline inertia and the fact we're accelerating. If I switch the simulator to steady state, there's no difference in WHP between Shoot_2R and Shoot_6F, as you'd expect.
The rest of the difference in FHP is due to slightly higher calculated losses in Shoot_6F.
So the estimated power loss in this case is 15% at that power & speed (103 kph).
Same atmos. conditions as the OP
This is the car producing 330 hp @ 5500 rpm in Shoot_2R, ramping at 100
280 WHP, 330 Fly.
I then changed to Shoot_6F, without touching the simulator :
282 WHP, 339 Fly.
The 2 hp higher WHP figure (<1%) is due to the assumed higher driveline inertia and the fact we're accelerating. If I switch the simulator to steady state, there's no difference in WHP between Shoot_2R and Shoot_6F, as you'd expect.
The rest of the difference in FHP is due to slightly higher calculated losses in Shoot_6F.
So the estimated power loss in this case is 15% at that power & speed (103 kph).
GreenV8S said:
I'm not involved myself, but I understand there are regulations about how the published figure is measured, including things like correcting to established environmental conditions and various tolerances for the accuracy of the measurement equipment. I've been told the manufacturers game the system quite extensively and will measure with equipment and under conditions which allow them to bias the results in their favour.
It's better than it used to be : the current standards basically require the engine to be tested "as fitted", so with all accessories, equivalent exhaust, etc.Back when we started this flywheel power malarkey, we kept getting low fly power figures for one particular model car, compared to the brochure figure. An unofficial chat with a couple of the manufacturer's engineers revealed that they couldn't get the official power figure on their engine dyno either...
but things are better now.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff