Inertia flywheel construction.

Joe T. · Feb 5, 2011

Every few years I get the urge to build an inertia dyno for testing engines. So it’s cold out and my brain is wandering. The flywheel for this dyno needs to have an adjustable mass to simulate different vehicles. This will work from the polar moment of inertia. Depending on the diameter, the flywheel could weigh over 1000#. The larger the diameter the less weight it will be but safety is a concern. This could be spun 10000 RPMs.
My theory is to mount a shaft on bearings for the axle. The mass would be something like 1” steel plates 16” in diameter. These would be added like dumbbells to set the mass for a given test.
Some of my questions are:
How large of a diameter could I spin 10000 RPMs without fear of explosions if the mass is steel? 7500 RPM would be common. Over that would not so much but possible.
What type of bearing would be best for support?
This is something that I may actually build in house. I want to keep it simple enough to build with a lathe, a vertical mill and a welder. I have ideas on the mass and a way to mount it and drive it. Basically I’m looking to find out what the limitations are to spinning a steel disc and bearings meant to hold up to this job.
All the data acquisition can be bought. I’m not concerned about that.

Thanks…

kurcules · Feb 5, 2011

I've kicked around the same thing off and on. One thing that occurred to me is that these days you don't need to actually sling around a half ton of iron to get the inertial effects you want: It would be much easier and more flexible to program a computer to simulate the effects of a large mass, but without the large mass.

Picture a much smaller fixed rotating mass with a fast brake (mechanical, maybe inductive), and torque and rpm sensors on the power input shaft. Throw in some software doing the appropriate math (very quickly) and you've got what you want - but much lighter, cheaper, safer, and more flexible.

HelicalCut · Feb 5, 2011

Steel is a very poor material for flywheels due to low strength to weight and the effect of the stress concentration caused by the hole in the middle. To get an idea of what a good design can achieve, steel turbine discs achieve a peripheral velocity of over 1000 ft/sec easily.

On the flip side a good carbon fibre design will go over 3000 ft sec in a vacuum.

The best inertial dyno is the car itseslf or a computer controlled load. If you are still keen on the flywheel, look for a scrap steam or gas turbine in the 7 - 10K rpm range and remove the buckets.

Ken hosford · Feb 6, 2011

10000 rpm 1000 lb fly wheel

warning the the imbalance forces go up with the sq of the speed at 10000 rpm it better be balance quite well . Or it will travel across the floor or take its bearings out quickly

motion guru · Feb 6, 2011

Kurkules has the right idea- I have designed a few chassis dyne's for testing automotive drivetrains. The use of a torque transducer and regenerative drive can make if fairly easy to simulate a small lightweight engine or car or a heavy V-10 or large truck.

Joe T. · Feb 6, 2011

I’m looking to avoid simulations and keep the computer for data logging only. The problem with simulations is they often need “facts” that aren’t always available for the math to reflect reality. Peoples “estimates” rarely help the cause. I’m looking to simplify things as well as only relying on a third party for data acquisition.

If the flywheel is made up from ½ - 1” thick plates in the 60# range, I should be able to balance each piece separately and I would hope very precisely. I would think this method should be very simple and repeatable. Not to mention, light enough to lift without killing myself.

I’m hoping to keep the overall length of the mass under 24”. This makes the diameter a big deal. How would I determine a maximum diameter for a steel disc turning 10000 RPM? Or what would be an ideal material for the flywheel other then steel?

motion guru · Feb 6, 2011

Joe,

admittedly, the cost of a high performance drive / motor to simulate a load is in the $30k range - the math for the simulation is the same as would be required to figure out the inertia of the steel plates required to simulate a drivetrain inertia - and there is a variable spring constant associated with the driveline and tires as well.

As far as calculating the stresses on the steel - while not trivial - certainly in the scope of a 2nd year mechanical engineering student. I bet entire junior year ME class labs have been tasked with this in Iran.

Seriously - if you want to make sure you don't create shrapnel . . . the second half of this document does a good job of laying out the design constraints.

http://nptel.iitm.ac.in/courses/IIT-MADRAS/Machine_Design_II/pdf/3_7.pdf

surplusjohn · Feb 6, 2011

I remember seeing the dymos at Bendix in Utica, these were in heavy concrete blockhouses with very thick steel plates for roofs so that when something blew the energy would be dispersed by lifting these very thick large steel plates up a few inches. very impressive!

dsergison · Feb 6, 2011

I think a steel disk of 16" diameter cannot be spun at 10k rpm.

look up surface speed limits of steel flywheels. the calculation is not hard but I don't know it offhand. but your # does sound about 2x the acceptable to me.

kurcules · Feb 6, 2011

You don't really need to spin it to 10K: You could double the rotational inertia and gear it down, for example. You're going to end up with a lot of watt-secs of energy in your flywheel, that's what you need to start your calcs on. And BTW you have to what we call "desaturate" the wheel after the test: you'll want a big brake on it to get it spun down after your test.

What you're going to find out is that your approach isn't very practical because of the energy levels involved. Take another look at the idea of simulating the inertial effects using a normal brake dyno as a basis. Heck, I don't know, but I suspect some modern dynos can already to that. Or if they can't, there's a product/add-on waiting to be built...

jim fuchs · Feb 6, 2011

try here for dyno help

TDK Motorsports Inertia Dyno Testing

pistonskirt · Feb 6, 2011

Sat on the couch on Sunday night I dont have the figures to hand but our Dynojet inertia dyno's drum is roughly 18" x 18" and around 1000lbs, it probably runs to a maximum of about 4000rpm & I certainly would not like to be around it spinning much higher :crazy:

If I understand your post correctly you are thinking of running the flywheel direct coupled at crankshaft speed, in which case even if you could find something capable of withstanding 10,000rpm starting & warming up the engine would be immensely difficult. If indeed you did get it spinning that fast you would then need one hell of a braking system to stop it safely, I can assure you that if an engine were to throw a rod you will want to stop it FAST.....before it saws off a couple of cylinders & throws them through the wall.
Any rotating brake parts will have to be included in the mass calculations & of course be capable of withstanding the rpm's.

Methinks you need at least a 2 to 1 drive reduction & a clutch sytem to make things not only safe & affordable but useable.

regards

Brian

Joe T. · Feb 6, 2011

Looking up some formulas in my Machinery's Handbook, looks like I'm asking for trouble. Based on 60,000 psi tensile strength 16" od 1"-ish thick the safe rpm is only just over 6100.

This seemed like a very basic formula but since the #'s aren't even close to what I need I think I'll put this back on the shelf.

S_W_Bausch · Feb 6, 2011

Wikipedia has a good entry on this subject [don't ask me to spell, uh, dynanometer].

This caught my eye:

Hydraulic dynos are renowned for having the absolute quickest load change ability, just slightly surpassing the eddy current absorbers. The downside is that they require large quantities of hot oil under high pressure and the requirement for an oil reservoir.

RDL · Feb 6, 2011

Joe T. said:
... Depending on the diameter, the flywheel could weigh over 1000#. The larger the diameter the less weight it will be but safety is a concern. This could be spun 10000 RPMs. ...
All the data acquisition can be bought. I’m not concerned about that.

Thanks…

How are you going to balance the flywheel? Static balance will not be sufficient and two plane balancing will require specialized equipment which is not cheap. Most software based balancing systems will not have the filters you will need to isolate the noise of those loaded bearings. Have you made a decision about your bearings?

Raymond

tailstock · Feb 6, 2011

Get a variable pitch torque converter like is used on a big truck dyno brake system. Cat had them up to about a thousand hp.

HelicalCut · Feb 6, 2011

Simulations are actually much easier, the reliance on facts is no problem, they are easy to obtain. The vehicle mass is known the ratio of the driveline is known and the polar moment of inertia of the engine can be found in the blip of a throttle. Computer controlled loads can even simulate traction and drag. A real inertial element, is only valid one design point meaning one gear in a particular mass of vehicle.

The only requirement is a fast load control A few disc brakes on a shaft with forced cooling and a hydraulic system would probably do. If four 350mm disc brakes can pull 1500Kg of vehicle to a stop in half the time it took 600Hp to get it there then such a setup might be what you are looking for.

Discs are bad shape for a flywheel, that is why they have low maximum rpm, there is a constant stress profile used for most turbine discs which allows about 50% more speed than a plain disk. In a disk the hole for the shaft is a stress raiser which reduces the allowable stress by 50%

delfi87 · Aug 18, 2019

Have you ever created flywheels for cars like this?

john.k · Aug 18, 2019

You can have a dozen or twenty smaller diameter discs,not necessarily just one big one....in fact,the loading could be varied by removing or adding discs.

Milland · Aug 18, 2019

Reported as spam. But thanks for bringing the thread back, I'm glad that guy didn't kill himself with his dyno design.

Inertia flywheel construction.

Cast Iron

Cast Iron

Stainless

Cast Iron

Diamond

Cast Iron

Diamond

Diamond

Diamond

Cast Iron

Cast Iron

Hot Rolled

Cast Iron

Diamond

Stainless

Hot Rolled

Stainless

Plastic

Diamond

Diamond

Similar threads