3 ms·

Seems to me they haven't built a clock: they've built a tricorder. Put an array of these in a box. Then computationally map the changes in gravity from each cl

Seems to me they haven't built a clock: they've built a tricorder. Put an array of these in a box. Then computationally map the changes in gravity from each clock. Bingo presto, you've just created a 3-D model of the mass in the local area.

pavel_lishin 12y ago
http://en.wikipedia.org/wiki/Robert_L._Forward#Forward_Mass_Detector http://en.wikipedia.org/wiki/Robert_L._Forward#Forward_Mass_...
- trhway 12y ago
  it would take some time to develop "gravitationally stealth" F-22 :)
  - pavel_lishin 12y ago
    Can't wait to invade a country to secure America's source of cavorite.
jofer 12y ago
For what it's worth, just directly measuring changes in gravitational acceleration is much easier and much more sensitive to changes in mass distribution. We've been able to measure gravitational acceleration very precisely for over a century now. An 80-year-old LaCoste & Romberg gravimeter will do the job quite nicely, though it's slow to use and you need to know the elevation you're taking the measurement at very precisely. (Interestingly, the ones made before ~1950 are more precise than the ones made in the 60's to 90's. It's basically a very well made and well calibrated spring. When they switched to mass-producing them, the quality fell.) Now you have a second problem, though... Regardless of whether you've measured things through gravitation acceleration or time dilation, going from the measured effect back to an actual mass distribution is a non-unique inverse problem. There are an infinite number of equally correct solutions (and a larger infinite number of incorrect ones). You can make a pretty good guess at what the mass distribution by applying reasonable a-priori constraints, but there's no single unique way to get the mass distribution from the gravitational effect of the mass.
ISL 12y ago
I build precision gravity sensors for a living. We tried doing something similar and demonstrated a proof of principle. It works, but gravity is often unkind to the experimentalist, chiefly through the combined facts that many mass configurations can yield the same signal and that the inverse-square law requires close proximity for good resolution. Almost any other sensor system you could dream up is more efficient. For three-dimensional arbitrary mass distributions, the inverse problem is extremely hard. For 2-D imaging or simply localizing a mass, things are easier. If there's any application for pure gravitational sensing that can resolve the position of a higher-density mass to a few centimeters over distances of perhaps a meter or that can make simple statements about a meter-scale mass distribution, please drop me an email. We've tabled the project because we don't know of a single use for it, academic or commercial. Device cost would be in the low hundreds of thousands of dollars and require careful operation. We've thought hard about this, but haven't ever found an application for which some other sensor wouldn't be far more appropriate. X-rays, neutrons, resistivity, clever weighing, optical techniques, microwaves, touch probes, three-year-olds, lemurs, optical imaging, you name it, it's probably cheaper, better, and faster.
- jofer 12y ago
  If you can improve on current gravimeters, it should be marketable. Potential fields surveys are still a big business with a lot of customers. (As an exploration geologist in the oil industry, I use gravimetry on a daily basis.) Generally speaking, though, accuracy of the sensor is rarely the limiting factor in gravity surveys. For land-surveys, it's precisely knowing your elevation, correcting for the "unwanted" mass distribution around you, etc. For mobile surveys, it's correcting for the acceleration of whatever vessel the instrument is on. Any ideas you might have for improving the state of the art for the mobile case would probably be _very_ marketable. On a separate note, though, the inverse problem, while fundementally very non-unique, is still solvable for many practical problems (e.g. we know the range of density of the materials involved and we can make a reasonable starting guess for the distribution of mass). Regardless, you're usually interested in distinguishing between a few scenarios that can be easily forward-modeled.
  - ISL 12y ago

pavel_lishin 12y ago

http://en.wikipedia.org/wiki/Robert_L._Forward#Forward_Mass_Detector http://en.wikipedia.org/wiki/Robert_L._Forward#Forward_Mass_...

trhway 12y ago
it would take some time to develop "gravitationally stealth" F-22 :)
- pavel_lishin 12y ago
  Can't wait to invade a country to secure America's source of cavorite.

jofer 12y ago

For what it's worth, just directly measuring changes in gravitational acceleration is much easier and much more sensitive to changes in mass distribution. We've been able to measure gravitational acceleration very precisely for over a century now. An 80-year-old LaCoste & Romberg gravimeter will do the job quite nicely, though it's slow to use and you need to know the elevation you're taking the measurement at very precisely. (Interestingly, the ones made before ~1950 are more precise than the ones made in the 60's to 90's. It's basically a very well made and well calibrated spring. When they switched to mass-producing them, the quality fell.) Now you have a second problem, though... Regardless of whether you've measured things through gravitation acceleration or time dilation, going from the measured effect back to an actual mass distribution is a non-unique inverse problem. There are an infinite number of equally correct solutions (and a larger infinite number of incorrect ones). You can make a pretty good guess at what the mass distribution by applying reasonable a-priori constraints, but there's no single unique way to get the mass distribution from the gravitational effect of the mass.

ISL 12y ago

I build precision gravity sensors for a living. We tried doing something similar and demonstrated a proof of principle. It works, but gravity is often unkind to the experimentalist, chiefly through the combined facts that many mass configurations can yield the same signal and that the inverse-square law requires close proximity for good resolution. Almost any other sensor system you could dream up is more efficient. For three-dimensional arbitrary mass distributions, the inverse problem is extremely hard. For 2-D imaging or simply localizing a mass, things are easier. If there's any application for pure gravitational sensing that can resolve the position of a higher-density mass to a few centimeters over distances of perhaps a meter or that can make simple statements about a meter-scale mass distribution, please drop me an email. We've tabled the project because we don't know of a single use for it, academic or commercial. Device cost would be in the low hundreds of thousands of dollars and require careful operation. We've thought hard about this, but haven't ever found an application for which some other sensor wouldn't be far more appropriate. X-rays, neutrons, resistivity, clever weighing, optical techniques, microwaves, touch probes, three-year-olds, lemurs, optical imaging, you name it, it's probably cheaper, better, and faster.

jofer 12y ago
If you can improve on current gravimeters, it should be marketable. Potential fields surveys are still a big business with a lot of customers. (As an exploration geologist in the oil industry, I use gravimetry on a daily basis.) Generally speaking, though, accuracy of the sensor is rarely the limiting factor in gravity surveys. For land-surveys, it's precisely knowing your elevation, correcting for the "unwanted" mass distribution around you, etc. For mobile surveys, it's correcting for the acceleration of whatever vessel the instrument is on. Any ideas you might have for improving the state of the art for the mobile case would probably be _very_ marketable. On a separate note, though, the inverse problem, while fundementally very non-unique, is still solvable for many practical problems (e.g. we know the range of density of the materials involved and we can make a reasonable starting guess for the distribution of mass). Regardless, you're usually interested in distinguishing between a few scenarios that can be easily forward-modeled.
- ISL 12y ago