VetNutJim

Just a comment on your non linear brake application.

My understanding is that drum brakes are usually "servo actuated" whereby the friction of the 'leading' brake pad helps actuate the trailing pad.

From the looks of your first dyno, this sems to be the case as there is a large pad and a smaller one.

You may be able to reduce this effect soewhat by reversing the physical location of the two brake pads.

BTW: Absolutely INGENOUS!

AustinT

Yes I have a computer, I am using the Innovate motorsports wideband as a datalogger. A signal conditioner between the loadcell and lm1 gives me a 0-5 volt output. The Cell is rated at 2500 lbs so 2500 lbs=5 volts. The lm1 is also plugged into the computer so I have realtime gauges of what is happening.

Ive thought about computer control but I dont know the best way to go about it.

Too bad I still dont have the cruise control functional on the car. I could set it at a rpm/mph and slowly apply more brake force. The car is a manual so it cant shift down. As soon as the engine could not hold the rpm for a given load I would know the max power for that rpm.
Does anybody know, will a cruise control reach WOT?


Hmm Ive never came across this. I know each wheel has 2 brakes. These brakes are the same size. The brakes pads are built where one side pivots. The other side has a s cam that pushes the pads apart and into the drum. The side with the s cam does all of the moveing while the other side just rotates. Before I bolted on the dyno wheels I pulled the drums off and applied a small amount of grease to the s can. I also greased all of the zerks for the brake system so it would move as free as possible. The s cam has an adbrupt lift. I think this is where most of the problem is. It takes a lot of force to get the rollers to start up the s cam. When the y finnaly move they move all at once. If I replaced the brake cans (air cylinders) with hydraulics or manualy activated I wouldnt have this problem

JoBy

AustinT, could you email me a log file? joby@joby.se

JoBy

Thanks for the file.

I managed to import the data into Excel and do sime magic XY plotting.



It is very interesting to see. The load cell scale is to the left. Crank hp and tourqe to the right. RPM at the bottom.

This is calculated with 180 pound preload and 15% losses.

You had a good loading point between 2000 and 2500 rpm and you see the yellow rising HP curve below. If you could do a WOT run and control the dyno brakes to let the engine rpm slowly rise, then you can plot the dyno result with Excel like this.

Grumpy

One thing you want to make sure of is that the car is square to the rollers. If not, then there is an incredible amount of HP wasted in flexing the sidewalls of the tires. I watched some pros say it doesn't matter, and then couldn't even get close to touching the sidewalls after the one pull. And amazingly enough the car was down on HP from the last pull when it was square to the rollers. The sidewalls felt about 20HP hot.

Cobra289

Is 15% average for the loss in a manual transmission drivetrain?
4th gear is 1-1 right?

I use an accelerometer to measure my torque and the related Power but on the road.

I was always being advised to make my runs in third gear. There is where the car produces the most peak power.
It goes like this:
Run the car in third gear to 1500-2000 rpm, you select the best rpm taking care that the car runs to WOT without hesitation.
Then press the throttle till the floor smoothly but quickly till RED line, at this point you depress the clutch, if you have one, and you let roll out.

Torque is a relationship between RPM and Power, I think that you should introduce the weather conditions in your calculations like; Pressure (mbar), Temp, Height, %RH (Humidity).

BHP meant "horsepower at the brochure."
The gross advertised BHP can be 40% higher than the real in the car installed with all the ancillaries connected. (A/C, Alternator, emissions controls and a full exhaust, etc).

SAE have the data how the horsepower should be measured.

One point to consider is: Calculate the radius of the tire correctly. Lest say you measure the circumferential length of the tire and divide by 2 times PI this will give you the real radius to calculate do that for tires, car and Dyno. You can rub a small amount of HP but just checking in a different way.
Regards,
Cobra289

Tom 400 CFI

onebinky

iTrader: (1)

Any updates for us yet?

steve8586iroc

How much more accurate is this than a g-tech pro? Other than not having to drive down a back road what is the real benefit to a chassis dyno? Not that I don't like chassis dynos or anything. Just curious.

Steve

AustinT

Havent had a chance to try it again. I am completly reworking the brake system. The air cannot be controlled. The axle assembly was unbolted from the ground and taken to the shop to work on. The aircans (aircylinder that apply the brakes) were removed to leave the slack adjuster. The slack adjuster is a 6 inch lever with a wormgear drive to adjust the brakes. The slack adjusters now have a 6 foot lever to control the brakes. One person will have to stay in the car and bring it up to WOT. The other person will stand in back of the car to control this lever. Since it will be 100% manually controlled I should be able to feather in the brakes and get a better dyno run. It will also have 12 times the travel as before for better control. I know its not the most elaborate way to apply the brakes but I belive it will work great. When Im all done I am putting together a webpage so others can do the same. Depending on what a person has access to, he or she should be able to build a dyno for 500-600 dollars. I have 235 into mine but I get free concrete.
Should have more time to work on it tomorrow, Ill keep you posted

I have never used a g tech but I hear they work great. The biggest pro that I can see for a chassis dyno is controlled loading. A g tech would be kind of like a intertia dyno. The car sits on top of two rollers and accelerates the rollers when the throttle is pressed. The power of the car is determined using the weight of the rollers and the change of rpm. Maybe someone else can clue us in on the benefits of both

JPrevost

Looks a bit leary to me but what works, works .
I like the load cell idea, simple and effective.
One idea might be to hookup a fluid coupler with pump to the pinion yoke to give you more reliable control of load. Basically a water-brake chassis dyno. What about adding water to the tires to increase inertia giving smoother control.
Also like RBob said, calibrate the load cell with a bunch of weights. Do this in the load cells intended temperature range.
Then to be as accurate as possible draw up a freebody diagram of the whole system to account for gravity and having the wheel forward on the dyno wheels.
I'm looking forward to your progress, keep up the good work.

AustinT

A water brake would be ideal. It would not be hard to couple the brake to the pinion. The trick is finding a water brake. Ive looked around and I cant find any prices on a water brake without buying a complete dyno. I dont know how much more weight the dyno tires would need. I would guess they weigh around 500 pnds together, the way they are.
Thanks JPrevost

BJM

Before you put the water brake on the pinion just realize that all of a sudden you will spend the next long while trying to decide how much loss there was in the gears. This would also imply you are giving up on the load cell idea which I think would work well once you get your brakes working.

JPrevost

How would it be giving up on the load cell? I'm not suggesting anything other than a better control of applying load. The load cell doesn't need to be moved to the brake housing, just keep it where it is and use the water brake as load only.

BJM

My mistake, never mind what I said. I was thinking of an engine brake where the torque is measured by how much a lever arm attached to the water brake housing pulls on a small load cell. I was thinking you would measure torque there too.

The water brake would make a great variable load. I have lost track if there is a differential gear on this axle you are using. You may want one and put the water brake on the differential pinion at a ratio of 3-4:1 which brings the speeds and torques in line with a normal engine brake which keeps it small. Doing it at the roller's slow speed and high torque would overwhelm a small water brake.

AustinT

The dyno axle does not have the pinion and ring anymore. I wanted to keep things simple so I removed the gears and pressed a piece of pipe between the axles. I could put it back together for a water brake but I dont believe it will be necessary. I finished welding the lever that will control the brakes. I also measured the ratio between the lever and the movement of the brake pads. For every 12 inches the lever moves the pads move 1/16" of a inch. I should be able to feather the brakes and get good dyno pulls. I agree a waterbrake would be the best but I dont know where to find one. The brakes that I have now will last for a long time if I dont glaze them over. On one of my first runs I did not have the brakes adjusted properly. One brake was doing all the work and I could smell it starting to get hot. I touched the drums to see which brake was hot and only one drum had heat in it. The dyno run lasted about 2 minutes one one brake. I would think it would last over 4 on two brakes

Z69

I'll try explaining this a different way..
Torque definition=(In Short) a force applied about a radius to cause rotation.
Since your cell isn't turning nor is it mounted on a radius. You are not measuring torque. Your measuring Force. So you can't calculate Hp as has been done above.
It's been about 17 yrs so I don't recall all the formulas you'll need to get to HP, but it can be done.
Force is equal to mass times acceleration or F=MA.
And IIRC you can use A to get to Power.
Sorry I can't get the rest for you. No time. It took 3 days for me to get time for this post .

As far as brake drag, just adjust the brakes 3/4 or one turn back and the shoes shouldn't be touching.
Also, if you are accelerating the whole time, some HP will be lost to acceleration. And your not measuring that. If you want a little more accuracy, try to hold a steady RPM. Or drag the motor down from high rpm with the brakes so that there are no acceleration losses.
Also engine oil temp and coolant temp will throw your repeatability off if they are not close going from test to test.

JoBy

That first dyno graph makes perfekt sense.

Look at the yellow, that is horsepower. Up is more and down is less. RPM is sideways, just like any dyno sheet.

When driving a car you don't use full power all the time. At 2000 rpm your engine might make 100 hp at full throttle. At anything else it will make less power.

On a 'normal' dyno run you only use full throttle, and the graph is cut to only show the full throttle run.

On this graph it is a lot of part throttle, slow increase and decrease of rpm.

That is the major problem with this dyno, Adjusting the brakes so you can make a nice full throttle run thru the RPM's.

On this the lower graph I cut out everything that you should not look at. The part that is left is a dyno run from 2000 to 2500 rpm.

JoBy

It is measuring the force at the radius of the wheel, so yes you can.

Z69

Duh,
Should have reread the posts again..

AustinT

Yeah, its kind of a trick getting the car on the dyno. I just threw a couple of logs between the dyno and concrete to bridge the gap. I started on a couple of ramps tonight using some 14 inch I beam. The wheels do have lug nuts. Semi wheels are a little different than car. The studs you see in the picture are the nuts. Another wheel fits on these studs and then the lugnuts.


The axle is again bolted down and ready to go. I really think it will work well this time. Thanks to all for your help and a special thanks to JoBy for the help with the numbers.

StungByStinger

Sweet!

JaST

Not if it uses intake vacuum to control throttle position (as most do). The available vacuum diminishes as you reach WOT, and on a dyno run, probably, even a vacuum accumulator would be depleted before the run was completed.

This concerns drum brakes being "self-actuating".
All internal drum brakes are self-actuating. There are 2 types, the names of which elude me at the moment, but here's a link to Howstuffworks.com concerning the effect. Basically, the shoe turns with the drum once activated, and it is then wedged into the drum. As HSW depicts it, the leading shoe forces the trailing shoe into the drum, creating rather dramatically higher braking capabilities than would be expericed by the hydraulics alone. I'm not following how your Semi brakes are setup, but if the two shoes pivot on the same end as they are actuated with the other side "floating", they work this way. The other way still generates more force than the actuator (hydraulic in the case of the automobile, but again, I'm not sure about the Semis) could produce by ramping the individual shoe into the drum. The shoes are relatively isolated, but the leading shoe will brake much "harder" than it would if only the actuation forces were pushing it into the drum. I believe this is the style used in your setup.
Obviously, this is normally a very good thing, as you get "more brake" for your "actuator buck", but one of the drawbacks is that the linearity of drum brakes is difficult to engineer. Grabbing or "sudden actuation" is a known hurdle for builders of drum brakes.
In other words, I'm still reading back info to find out why you scratched the generator or hydrobrake idea, but drum brakes are probably going to be VERY difficult to make consistant.
Here's a link to how I presume the airbrakes you have work. There's also slight disucssion of a couple of other possible drawbacks to using drum brakes...

Hope I've been helpful.

WesFlorida

How about bringing the engine to the max speed you want to measure to, say 5500 rpm and then flooring it and applying the brakes untill the engine has been slowed to 2000 rpm or so, logging the whole thing.

Low C1500

All of your rotating assembly would increase the load cell values. So instead of losing 30% of your power through the drive train, you would be gaining 30%, because the rotating stuff has momentum

JoBy

Yes, The best measurement would be to start a WOT run at 2000 RPM and let the RPM rise to a max like 5500 RPM. Then the dyno would apply more brakes and force the engine back to 2000 RPM, still WOT.

If the acceleration and deacceleraton was the same rate, then you can average out the momentum from the rotating mass.

Low C1500

WOuld be really neat if one could make this work, then You could have a good engine power value, but of course there is always the loss friction.

onebinky

iTrader: (1)

S-Cam and Wedge type

AustinT

This is the s cam type. I have a few wedge brakes on some axles and I dont like them as well as the s cam. The biggest drawback to the brake lever is I need someone else to help me run it. The axle is all ready to go if I can find a second person

Tom 400 CFI

I live in Park City. Where is Cannonville?? I wouldn't mind helping at all. I find this project of yours extremely interesting.

-Tom

Lionsden

I agree with Tom.

AustinT

Cannonville Is in South Utah by Bryce Canyon. It is around 5 hours from Park City. The Dyno Is still in the test stage, I would hate for you to drive 5 hours and be dissapointed. But sure, if your down this way let me Know, Thanks


Tried the new brake lever system and it seems to work great. I did not have the load cell on the car during the pull so I dont have any numbers. Watching the rpm guage in the car while a friend slowly applied the brakes, it looks like it can be adjusted and held in 300-400 rpm increments. Recalibrated the loadcell and it looks like it was off slightly. The loadcell was calibrated with 633 pounds but I increased the calibration to 1500 pounds for a more accurate calibration.
It took awhile to get the calibration set. Somtime during the calibration I accidently screwed in the eyebolts. The bolts pushed against the center off the loadcell and simulated a 300 pound load. I tried and tried to calibrate the cell but zero calibration voltate was too much because of the preload. Finnaly after a few hours I noticed the bolts.

Tom 400 CFI

Gotcha. Now I know exactly where you are. I would consider coming down there. Make a weekend of it. If you ever want to run your car at RMR, I'll give you a place to stay for the night/weekend or whatever.

Hi Dan! Good to "hear" from ya!

SuperMag

Newb chiming in here. I think I know why your values are so low...

The only thing the loadcell is measuring is the force levered into the earth...

You also need to figure out how much heat you are generating with those truck brakes. After all, it takes power to make heat. Lots of it....

Low C1500

Most of the aplied force here is getting turend into heat via friction. Shouldn't matter though about the brakes, as the pull on the load cell (energy) is just being transfered to the brakes, not lost.

Now if he was to aply his own brakes to the rear tires, this would be a loss, not a transfer, in terms of the pull on the load cell.

SuperMag

Uh-uh. Look at the pictures again... the power from the rear wheels is going to two places: (1) through the load cell and into the earth via that big steel pole (which is the only force the loadcell is measuring), and (2) as frictional losses generating heat in the truck brakes in the "dyno."

If our man were to figure out how many BTUs of heat were being generated by those truck brakes, he could convert BTUs to HP (a simple formula), add that to the HP figure derived from the loadcell reading, and multiply the resulting figure by 1.15 (frictional losses in the drivelines), he'd be spot-on.

The trick is going to be measuring (estimating) the total heat output of the brakes...

AustinT

Had to think on the brake heat. The brakes are burning off the horsepower of the engine and creating a load on the vehicle. The load is measured via the loadcell. The amount of heat made in the brakes is equal to pull at the load cell.

Low C1500

You need to look at it as follows:

The more friction that the brakes make, the more force is required by the car to turn the axle, now the load cell reads this force by pulling equally as much in the opposite direction as the car.

Also the load cell isn't dissapating any energy, its only reading the transfer off car power, into heat (most of the dissapation) and truck axle acceleration (the lesser of the two).

Now the contact point of the two sets of tires is a different story.

AustinT

That would be great. I built the car to play around with at the strip but live so far away I never have ran it. Ill keep in touch with you for an open weekend, Thanks




More Dyno numbers tonight, JoBy would it Be Ok if I sent you a couple of more graphs. I know how to download the info into excel but I dont know how you made the graph. The dyno worked so well tonight I am still smiling. The load can be controlled easily. The only reason to go to waterbrake would be longer run times. After a 3 min pull I checked the temp of the dyno brake drums with a infared Thermometer, 500 degrees F.
For what I am doing the brakes will work fine. The car was centered on the dyno so it had little preload, only 24 pounds.


Here are some WOT numbers preload compensated, 4740 rpm at 890 pounds of pull at a AFR of 10.32

Numbers from a 2nd run preload compensated. 3790 rpm, 904 lbs of pull at 10.26 afr


The Dyno runs tonight are more accurate than last week because of the new way of applying the brakes .


Looks like the AFR is a little rich. Would there be substancial gains to lean it out a little? Any way to log the spark timing?
Thanks again everbody

AustinT

JoBy Could you please explain on the 15 percent loss.

Following your numbers on tonights numbers I get 246 corrected Hp Is this right?

onebinky

iTrader: (1)

I think he's referring to the 15% parasitic loss typically seen with a manual gearbox drivetrain between the flywheel and tires.

JoBy

If you have 15% friction loss from the engine to the rear wheels then you have 85% of the power left.

When calculating crank hp from rear wheel hp you get another percentage.

100/85 = 1.1765. ( or 17.65% more at the crank.)

17.65% of 85 hp is 15 hp.
15% of 100 hp is 15 hp.

Do you have another log file I can look at?

omahaturbo

If you wanted to dyno future cars and not have to mess with figuring out the car's gear ratios or worry about what gear you are in everytime you could put a trigger on the trailer wheels to log the trailer wheel's rpm along with the engines rpm. Then you could use the ratio of the 2 rpms at given point with the trailer wheel's radius to calculate the engines torque and then hp. You could then put the formula in excel so that it would work for any car in any gear without having to worry about any more ratios or tire size as long as you logged both rpms. Just log both rpms and the force on any car, paste into a premade excel template and instant dyno dumbers without anymore math, just like a real dyno. You could also see if there is any clutch or tire slippage if the ratio of the 2 rpms is not a linear line, just like on other dynos.

JoBy

Actually, before reading this thread I and a couple of my friends had a similar idea of using a truck rear axle to build a dyno.

By mounting the axle above a pivot point the axle can swing a bit back and forward. It is held in place with an oil filled hydraulic piston. Instead of using a load cell the oil pressure in the piston is used to measure the force.

Ánother benefit is that you can tie down the vehicle any way you want without interfering with the measurement.

omahaturbo, You got the next part. By measuring the rpm and force on the dyno axle hp can be calculated regardless of vehicle tire size and gear ratios. By also measuring the engine rpm the hp numbers can be transferd to that rpm.

Then we want to use an hydraulic retarder as load, but using the brakes would also work.

AustinT

Thats a good Idea. Shouldnt be very hard to do I could use a sensor and frequency to voltage circuit to do the job. The innovate motorsports software that I am using allows me to scale the numbers.


My axle was perfect for your application before I took the torch to it. I had some plans alot like the plans you posted but the axle was in a bearing so it would spin. A arm was going to be welded off the axle for a load cell. The reason I went with the method I now have is I could not find any bearings large enough for the axle. I worked in a aluminum and bronze casting factory earlier this year. I was going to cast some bronze bearings and add some zerks. Finnaly I talked my self out of it,I didnt think I could make a near friction free bearing. I should have known I could put the pivot away from the axle and use the axle as the swing arm. Look for a walking beam type suspension from a semi. the axle already has an arm attached to it. My axle was this way but I cut off the beam becuase it was in the way. What do you think about using a loadcell in place of the hydraulic ram readout on your dyno? I can send you the info where I bought the loadcell and signal convertor. The loadcell was only 135 and signal conditioner was 85. The signal conditioner I orderd is for 0-5 volt. You can order it with 0-10v and 4-20ma or rs232.


What kind of hydraulic retarder did you want to use? Do you think it will burn off the Hp it needs to? I have another post hear at thirdgen from a couple of months ago before I stared the dyno, maybe you saw it? This was the biggest concern before I started the project, its difficult to burn off the horses. I thought of generators, hydraulic pumps with a needle valve and a few other ways. I called land and sea today and priced a waterbrake, 6000. I think Ill stay with the drum brakes for awhile. Did you get the last dyno runs I sent today? The dyno worked great and can be controlled in small increments. The only problem we had is the guy who ran the lever system behind the car could not tell if I was WOT or not. He had the laptop so he could see AFR, rpm and load but he did not know how far I was into the throttle. When He applied the load I would slowly add more throttle so the rpms would stay at 4000. He thought I was at WOT so he applied the brakes very slow. The first minute and half was spent bringing the car up to WOT and the dyno run lasted around 30 seconds. The second run I put my hand out the window and signaled when I was applying the throttle and when I reached WOT. We reached WOT quickly and then he began to feather the brakes. The run was more usefull becuase most of it was WOT. Im going to use a TPS sensor so he can see how close I am to WOT.

JoBy

The loadcell would work but you need to preload the cell like you do with the car.


I am from Sweden so there are plenty of SCANIA trucks around.
http://www.scania.com/products/truck...ania_retarder/

The retarder is a hydraulic brake with an oil to water heat exchanger. It is bolted to the rear of the transmission. There are many sizes and some are way too powerful.
No, I did not see it.

AustinT

I have been researching the hydraulic retarders and they look like they would work perfect. We own a small fleet of cement mixer trucks but they have jake brakes so I was not familiar with the retarders. I emailed for a retarder that would burn off 500 hp and they wanted 9000 for the complete kit, too much. For now I will stay with the brakes. I bought a linear actuator off ebay to run brake lever. It has a 11 inch stroke and takes 41 seconds to complete the stroke. It should make it easy to get reliable numbers. I have been thinking more on computer control for the brakes like you suggested. What if I bought one of the add on cruise controls like the one made by audiovox? Could I simply add some magnets to dyno wheels for the sensor pickup? I think I might be able to put something together to run the linear actuator with the add on cruise control.

JohnL

You can't take moments about the tyre-to-tyre interface because it's not in equilibrium there. You need to take moments about the truck axle. You then have the braking torque around the truck axle cancelling out the engine's torque around the 3rdGen axle, (while the rear axle load and the truck axle reaction pass through the same centreline ideally), and the calculation becomes (modifying JoBy's earlier format):

1050-180 = 870 lbs at 3900 rpm

245-45-17 => 12.84" radius = 1.07 feet, PLUS 445-50-22.5 (assuming you've using something like Bridgestone super singles) => 20.0" radius = 1.67 feet = 2.74 feet total lever arm.

870 lbs * 2.74 ft = 2382 lbs ft at the wheels.

2382/ 4.10 = 580.9 lbs ft at the crank.

580.9 lbs ft * 3900 rpm / 5252 = 430 hp

Corrected with 15% loss. ( 1/0.85 = 1.1765 )

580.9 * 1.1765 = 683 lbs ft

430 * 1.1765 = 505 hp

I think these numbers are high because your 3rdGen wheel is in front of the truck axle centreline, and this is increasing the load on the load cell. And/Or I've used too big a truck wheel diameter in my calcs.

John

AustinT

Here are some more reliable numbers than the first dyno run. Also the car tires were centered on the dyno tires. The only preload was 25 pounds, probably just the weight to hold the chain up. Dyno tires are 425/65r/22.5

JoBy

JohnL,

Both axles see the same force that the load cell measures.
The radius is different and the force is the same. The bigger dyno tires will have more tourqe but less RPM. The horsepower on both axles will be the same. Knowing the RPM, force and radius on one of the axles is enough to calculate the horsepower.

The radius on both axles should not be added.

245-45-17 => 12.84" radius = 1.07 feet.

890 lbs * 1.07 ft = 952.3 lbs ft at the wheels.

To make the calculation a bit different I calculate the tire RPM.
Tire RPM = Engine RPM ( 4740 ) / drivetrain ratio ( 4.10 ) = 1156 RPM.

RWHP = 952.3 lbs ft * 1156 rpm / 5252 = 209.6 hp

Corrected with 15% loss. ( 1/0.85 = 1.1765 )

232.3 * 1.1765 = 273.3 lbs ft

209.6 * 1.1765 = 246.6 hp