## Applied ME: Dynamic load calculation, drag force correlation. Load cells and LabVIEW?

Hello,
First time post long time lurker! I've got an interesting project I am working on which involves a few MechE subjects. The overall goal (for now) is to determine the drag force of a cylindrical cone headed object in sand under various pressures. The movement of the cylinder is the tricky part.

The subproblem is to measure the impulsive force imparted onto the cylindrical object. I have a simple experimental setup in mind (maybe you can tell me otherwise :) ) It involves mounting a load cell on the end of a pendulum and aligning the vertical at rest position of the striking end of the cell with some kind of dowel pole connected to the cylinder submerged in my sandy test bed. setting the pendulum to fall at larger angles will direct more force.

One of the main problems I am having with connecting the student/on paper world to the real world is the measurement mechanism. I know load cells are usually used for static loads, so maybe the response will not be fast enough to measure this .2 second collision. Would an accelerometer be more suited? also, how do I actually collect the data? I am learning more about labVIEW and it seems like a likely candidate. Finally, an I havent thought this out very well, would these force measurements alone provide what we need to determine the drag on the cylinder? (given densities, pressures, displacements, velocities)
 PhysOrg.com science news on PhysOrg.com >> Galaxies fed by funnels of fuel>> The better to see you with: Scientists build record-setting metamaterial flat lens>> Google eyes emerging markets networks
 I don't really understand your experimental set up, can you post a drawing? Is the cone striking the bed of sand vertically? Load cells + dynamic forces = failure. Definitely go with an accelerometer. I'm not a big fan of labview. I have yet to use a labview program that didn't contain at least 10 bugs and I have used a LOT of labview programs. Although the highest version was LabView 6 so maybe it's gotten better. I don't know if you have enough info to determine the drag because I don't understand your experiment set up very well. Is the only force moving the cone through the sand the cone's inertia?
 Ok, sorry for the delay. I drew up a quick sketch in paint to show what I am trying to do. The sand box would just be a plastic bin with a small hole for the rod (which is connected to the probe) to stick out. I know there would be drag on the rod, that would be minimized during design. The probe would probably be much closer to the wall of the box than pictured. I forgot an axis but I think you can figure out how it would work by the way the pendulum is oriented.

## Applied ME: Dynamic load calculation, drag force correlation. Load cells and LabVIEW?

Strain gauge load cells are typically used for static measurements. Piezoelectric load cells are often used for dynamic experiments and are well suited to the type of experiment you're carrying out. Alternatively, an accelerometer rigidly mounted to the back of the head wouldn't be too bad, although the force range you can measure is then determined by the effective mass that the accelerometer is attached to. For a pendulum this will require a few assumptions and calculations. Personally I'd stick to the load cell. Problem is, they are generally a fair amount more expensive than comparable accelerometers.

There are two types of piezo load cells - externally amplified and integrated signals (you have to integrate the charge output to obtain a voltage varying with applied load - calibration data given generally in pC/N); and internally amplified signals (you supply an excitation voltage and a voltage proportional to load is returned - calibration data given in mV/N and referred to as IEPE, ICP or Piezotron depending on manufacturer). Piezo accelerometers also exist of the same type.

Either way you will need the relevant amplifier and data acquisition hardware. You can either amplify externally and then measure voltage, or National Instruments offer a series of integrated excitation / acquisition units such as the NI 9233 which can be run from a PC via USB.

Software wise you will need to either configure SignalExpress to obtain measurements or (better if you're a student with any prior experience of LabVIEW) use examples in obtaining load cell / accelerometer data to help you create your own VI in LabVIEW. As for bugginess of LabVIEW VIs, that's down to the individual programmer's code. I've spent the last two years writing standalone applications and development VIs and understand it enough not to cause too many crashes. My biggest complaint is about the installation size for runtime files and hardware drivers.
 Thanks timmay, you have provided some good information. Do you have any recommendations on where i could source this hardware? (ie accelerometer, piezo load cell) Omega is pretty expensive and more industrial than what I am looking for. Jameco has one for $60. Do anyone know of any other websites?  EBay is occasionally helpful, you might be able to pick up accelerometers there from time to time. A variety of companies that spring to mind are PCB, Endevco, Dytran...unfortunately, they will all be pretty expensive for the range of loads you are probably after. You'll also need the data acquisition and amplifying hardware. If you're a student at a university with a decent Mech Eng facility there is bound to be somebody who has the necessary equipment. That would be your best option as the accelerations and loads are probably going to be much higher than the 2g that the Jameco product can handle. That would be able to measure a striking force of twice the weight of the impactor, which will be pretty small. For instance, the maximum force allowable in testing of soccer shinguards is 2000 N, which is approximately 200g for the 1kg impactor. If you're a student at school you will probably struggle with the budget. In which case, try contacting your local university. You are most definitely on the right track with your initial ideas though.  Timmay you have given me some good info which is leading me in the right direction. Ive done more research and have decided to go with an impulse hammer and accelerometer combination. Currently I'm pricing everything out from equipment to DAQ and software. This is an academic project and my professor informed me that$3000 to run this experiment is "rather cheap." :P Its looking like a 100lbf impulse hammer coupled with a 500g accelerometer
 does anyone have an opinion whether a 500g accelerometer is overkill for this application? Should I opt for the more sensitive 100g?
 Sorry for the delay, Easter holidays and all that. There are two considerations here. One is the measurement error / noise levels due to the accelerometer's sensitivity, and the other is maximum acceleration measureable. You could try estimating the accelerations you will be measuring, perhaps by estimating the impact time and considering the change in the hammer's momentum at the impact, converting back from force. From experience it's nice to have as sensitive a tranduscer as possible, hence the 100 g accelerometer (if the accelerations are well within that figure) would be nice. But (again from experience) you may find that you wish to carry out impacts at higher speeds or higher energies, meaning that 100 g limit is capping the experiments you can carry out. If you can, I would recommend estimating those accelerations and then using information from the datasheets for each device, figure out what kind of errors you'll obtain. If the 500 g accelerometer isn't too dissimilar to the 100 g, get that as you can always use it to hit things harder and faster later on ;) Finally, when it comes to data acquisition, two important things to consider are acquisition rate and quantisation levels. Make sure the acquisition unit is fast enough to give you high resolution time-wise (the minimum I now use for typical 5 m/s impact tests is between 10 kHz and 50 kHz, preferably the latter) and has a decent analogue to digital encoder to give you high resolution amplitude-wise. I either use a 12-bit digital storage oscilloscope which requires you to specifically set up the range of voltages you're reading to get good amplitude resolution, but has a high acquisition rate, or a 24-bit National Instruments module which requires no adjustment of input ranges for smooth data but is limited to about 50 kHz. Let me know how you get on with system selection.