Accelerometer for crude absolute gravimeter?

[billybowden]

I'll preface this by saying this is a project just for fun.

Background: As most people would be aware, gravity decreases from the poles to the equator, or rather centrifugal "force" increases. I did an experiment weighing a 500 gram calibration weight at 38°S and 28°S and it was 0.4 grams lighter at 28°S. This was an OK result but it has many variables, such as accuracy of the scales or if the calibration of the scales changed between readings.

Question: I was wondering if it would be possible to build a crude absolute gravimeter using Arduino parts. This is not my area of expertise but I imagined using a 3-axis accelerometer and getting 2 of the axes as close to zero as possible, and the third would be the signal axis. However, looking at the many available accelerometers, I wasn't sure if a) this would give me the results I desired and b) which accelerometer to choose. From my initial research, I figured that sensitivity would be the main thing and I'd want it to be as sensitive as possible around 1g (gravity, not gram). The changes I'm trying to measure are about 0.01% per degree of latitude in the middle latitudes but if I could measure a change of 0.1% for 10° of latitude, that might still be interesting if it was repeatable.

I did find one related project, "Ardunio interface to measure Gravity Changes using a Gravimeter" but that was for a relative gravimeter, not absolute.

Does anyone think this would work and/or can offer some advice as to which parts to buy?

[jhaine]

Why don't you look at the specs of the readily available sensors such as those from Bosch?

Accelerometer BMA422

That gives specs for sensitivity etc from which you can do your own calculations. I think these things give you acceleration along all 3 axes so a quick bit of pythagoras removes the need to get 2 axes close to zero.

[jremington]

The Earth is not a sphere, which also causes variations in g with latitude.

[billybowden]

Thanks very much. The spec sheet for that part was much more detailed than others I've found.

[billybowden]

Good point. The variation in gravity due to Earth being an ellipsoid is about half of that due to the centrifugal effect (at the equator) so it is significant.