Practical load calculation method 1

[JeanMicheling]

Hello everyone,

I have a special application with a spindle that is planing wood. I'm looking for a way to calculate the load on bearings in operation. Since there are too many variables, it's almost impossible to have a good approximation by theorical calculation. Thus, I'm looking for a practical way to measure the load. I thought I could have a good value by measuring the deflection of the shaft in operation and then, evaluate the bearing load. My concern is about the way to calculate that value. Should I only compute the bearing deflection or the shaft deflection or both. The spindle has a pair of angular contact bearing 15 deg 100mm-150mm ligh preload at each end. Is there a general thumb rule for that? Will it be a good approximation to only calculate the bearing radial deflection?

Thanks,

 

[Skogsgurra]

Hi,

I think that deflection measurement is way too uncertain. If I understand right, the planing means that you have a mostly tangential force on your tool and that it will be transferred directly to the spindle shaft.

A simple method would then be to measure active power used by the motor when idling and when it is working. The difference is the power used to remove chips from wood.

Using this information and knowing spindle speed as well as tool diameter will make a fairly accurate calculation of the actual force possible. At least with around +/-30 percent accuracy - and that is probably the best you can get without mounting the assembly on force transducers.




Gunnar Englund

http://www.gke.org

 

[JeanMicheling]

I tried this method before, and by investigating, I realised that the load is not only tangential because there is a gap between each board that can create a radial impact. Depending on the the area the tool has been struck, the load can varie because of the taper effect. So I'm really not sure taking the motor torque is a reliable approximation.

 

[Skogsgurra]

My thinking is that you need to design for a maximum load. I also think that it is possible to identify a minimum radius where the tool is cutting. Most wood tools cut at their perimeters, but still.

I have done these measurements/calculations on high-speed spindles where active power can be read directly from the drive. The air gaps will not mean any increased impact - rather the opposite. The tool will start hitting the upper corner and chip that away (low power), the next chip will need a little more power and so on until there is a series of chips with same power until end of plank. There are no extreme forces when hitting the first corner so you do not need to take that into account. The direction of the force will be somewhat different. But since these initial chips are small, you do not really need to bother about them.

Measuring shaft deflection or bearing deflection still seems (to me) to be much more error-prone. Has anyone done that with success? And calculated forces with any accuracy? I think that we are talking high-speed here, not a few hundred RPM.

Gunnar Englund

http://www.gke.org

 

[rnd2]

I think skoggs is right in that it is just too difficult.
Here is maybe a compromise. With the spindle running maximum, apply also the maximum operational load and somehow and record it. Turn the thing off and apply the same load but this time in static. Measure the deflection before and after the load.

 

[Skogsgurra]

This is getting quite interesting, rnd2.

Apply same load, but static. Would you do that by pressing a plank against the still-standing spindle? In that case, wouldn't the force used to press plank against spindle be the answer?

Or have I got the configuration wrong?

Gunnar Englund

http://www.gke.org

 

[JeanMicheling]

Skog I would agree with you that the first cut will be low power one as far as the knives are brand new. As the knives start to wear, mechanics will sharpen them. That will create a roundness effect and the tool shape will start to look like a whole round wheel which will lead to radial impacts. The amplitude of those impacts will varie depending on the contact point.

rnd2 I think your idea can work. I should try it in a few weeks and let you guys know.

Thanks,

 

[rnd2]

skoggs Yes, pretty much. The idea to replicate in static mode the maximum force the device might need to operate. I thought it might be easier and safer to measure the force as well as deflection with the spindle stationary.

 

[Skogsgurra]

Jean,

Looking forward to your results.

Gunnar Englund

http://www.gke.org

 

[JeanMicheling]

Ok so I've just come back from the mill where I did my tests. Here are the points that must be payed attention to :

-Using a dial indicator with the right range. Too precise leads to a difficulty to take the readings. Not enough leads to the incapacity to take the readings.

-Forecast a stand to hold the indicator. In fact, the vibration on the machine false the readings.

-For mor precision, a good tip would be to film the indicator in operation. Therefore, it will be easier to exactly see what is the measurement.

-Try different angles on the shaft to make sure you have the optimum deflection angle.

This method works good and I would recommand it to anyone who need to get the loads on his machine and have no other to do it.

 

[Skogsgurra]

Jean,

Interesting to have your feedback.

I have been using LVDTs for similar measurements. They usually come with electronic modules where you can filter out the vibrations and get a steady mean value. They also allow you to select ranges - often between 0.01 mm to 10 mm - and that should be sufficient in most cases.

Gunnar Englund

http://www.gke.org

 

[JeanMicheling]

Skog, would you tell me what is a lvdt?

 

[Skogsgurra]

Sorry, it is a Linear Variable Differential Transformer. A rugged device used to measure displacement with great accuracy. See

http://www.sensotec.com/lvdt.asp

for some examples.

Gunnar Englund

http://www.gke.org