Data Extrapolation for a Hydraulic Damper

Through it in Excel, chart it, toss in a polynomial trendline, show the equation, and call it a day.

can u pls explain the procedure in detail?

1. Enter data in columns (force in column 'A', time in column 'B')
2. Highlight the data
3. Insert Chart - choose scatter plot
4. Follow the chart wizard (make sure x-axis data is time, y-axis data is force)
5. The chart will appear on your sheet
6. Right click on a data point in the chart
   
7. Select "Add Trendline" - pick a higher order polynomial type
8. Right click on trendline and pick Show Equation
   

Your procedure is correct as long as you stay in validity domain of the curve. This not always the case and although the approach is OK the results are not. They correct from the mathematical point of view and wrong from the physical. Extrapolating is not only a procedure but an approximation of what happens. It is very dangerous to look at it only as a math problem.

In the case we deal with the maximal value PC wants is most probably not where he thinks by only extrapolations.

Like Edgar said - tinker with Excel. Something doesn't look right in the given data.

x-scale is 2.686-t. Not extrapolated. Won't be, by me, neither. Maybe you messed up your recording.

....on the other hand, I have a minute till the kettle boils. If the last 2 results are wrong, I'd run with 7304.012378

Doh, I 'll try again.

I'll understand it eventually. I enetered one value wrong, but it looks like a fit

I have the feeling - combining the spring problem you had with the look of the curve force=f(time) - that you work in the field of guns concept.

To continue the discussion I would like to know on which continent you are located since weapons development does not make part of CR4 activities with the exception of bows and other arrows propelling equipments for which we have the honor to have among us a true expert (purrr....).

The F=f(t) looks like the expansion curve along the shell travel in the barrel of a cannon. The fact you do not come to good results in your attempt to have the initial force (in fact pressure) can be explained.

No, these are the readings of a hudraulic damper used for absorbing the shocks of slowly descending load.

Actually I want to know the reaction force given by a damper when the load suddenly falls down, due to some failure. One thing is sure that the damper is going to fail. So, to prove it I am just trying to extrapolate the graph value to give a force value for the calculated value of time i.e. 0.02 sec.

I think this will also give me the impact value of free falling weight.

Hi,

I just directed by the comment #3 & got the below equation ...

Y=5003.5x^2-24406x+25191

Y= Force, x =Time...

When i substitute the value of 0.02 for "x" in the above eqn.., Y found to be 24744.88.. Just check whether the outcome is correct....

Hari

Please indicate the kind of damper since the extrapolation should be done on PHYSICAL basis and NOT only on the measured values.

This is, which ever device type, the reason for the possible error.

Has the damper a constant lamination opening? Does it have several metering orifices which are closed by piston position? Do you know where they are placed? Do you have any record of the load position correlated to the pressure/force measured values?

Do you have any other details about the damper?

It is imperative to make, for the extrapolation, a correct even approximated "model" of the damper if not since you are with last value very far away from the region where you want to work the errors can be extremely important.

Do you have any other details about the damper?

Does that really matter?

Actually I was having readings of 12 dampers, and readings of some dampers were really erratic (I sometimes hate quality people). So I simply took the average of each reading.

May be that was the mistake. But, then I selected one reading (out of 12) which is little o.k. I will be able to post it only after two days since I am going on holiday. Those reading appx fit a curve of 4th degree polynomial equation (on excel sheet). The result somehow matches with my calculated value.

If you concentrate on the curve alone without understanding the true function then you can extrapolate the wrong way.

I shall try to explain and if I will have the time show you the curve you can expect.

When the falling mass comes to a contact it has a velocity "vo" and the damper is at rest in its end position fully extended. If it is pereloaded (some are) then the piston is pressed with a force "Fo" against the stop.

The damper moving mass is small with respect to the falling one and can be in 1st assumption negelected. The pison will move inside at start with vo and compress the oil (or let say the fluid used in the damper) and due to a higher pressure the fluid will flow through the orifice(s). The mass losses progressively its speed but in the 1st phase the pressure still grows it will reach a maximal value function of initial speed, preload and size of orifices and after it will decrease. Your contact force is NOT the maximal value. The maximal value depends on the parameters of the damper and the initial velocity at contact and the mass and the nature of used fluid and some other factors less important.

So only extrapolating your measurements which are surely after the peak you will get value out of range.

Sorry to insist on the PHYSICS of what you analyse the curve is nothing if you do not know what can lead to it.

Data extrapolation for a damper is in my experiance not straight forward but the following may be of some help.

Force in a damper is proportional to the square of the velocity. (from your figures I assume that the velocity varies, the stroke being constant?).

I also assume the medium in the damper is oil?

You will have many nonlinear effects ( Oil temperature changes which affect viscosity, The vapour pressure of the medium used, the actual configuration of the orifie (length, edge sharpness etc., etc,.

If you wish me to go on let me know. I built test equipment for shock absorbers (Dampers) for the Auto Industry.

Regards

Arthur E. Firminger

There are differences if the damper is for industrial use. Usually in automotive damper the orifice is constant (if not adaptive) in the industrial design there are several orifices closed one after the other when the pistion moves. Due to that the proportionality between piston velocity and force (via pressure) is variable over the stroke.

force time

829.57 2.6

1659.15 1.74

2488.73 1.35

3318.3 1.26

4147.88 1.11

4977.46 1.08

7300.27 1.09

7967.94 1.02

I tried various fitting curves using excel, and 4th degree polynomial is fitting the best.

y = 40465x4 - 282475x³ + 718623x² - 795991x + 328141

Thanks for your explanation. But, you see gathering all these data is really a headache. So, I prefer to go an easy way.

From the above equation the froce for 0.019 sec is 313271.6617kgf. The damper as per as told the vendor can take only upto 9500kgf force. So, it's bound to fail.

Hi, pc!

Just as an observation, your results make no sense from two points of view. First, the solution you found seems extreme and outside of probability. Second, if you are dealing with fail-safes on OEM equipment, the manufacturer would have noticed that normal working loads and failure loads were bound to make the damper fail because of all the returned equipment and lawsuits arising from resulting damage, if for no other reason.

Have you been in contact with the OEM regarding your concern?

If your "failure rate" calculation of the damper applies more force than the manufacturer of the new piece had intended, you probably have ordered machinery too lightweight for the purpose you intend.

Try this site or one like it, and automatically regenerate your curve as a double-check against your own calculations:

http://xrjunque.nom.es/precis/polycalc.aspx

Mark

I tried to explain but without success apparently that the way you extrapolate is "mathematical" and not at all related to the physics of a damper. if in your profession you will continue this way there are some risks of "personal failure " not only hardware.

I sincerely doubt that the force peak will be so high. It could be possible to that the force peak will be higher than the indicated limit for the damper but not at all so high as your extrapolation gives as a result.

Can you give more input about the damper as design, piston area and if possible any input about the orifices. Is it a single or multiple orifices damper?

Could you give more measured points for the other checked dampers? It would important to see how the dispersion looks.

Use PM chanel as for the other problem.