Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Can someone tell me if GP-B has results about GR, one way or the other?

  1. Dec 11, 2007 #1

    rbj

    User Avatar

    I just can't figger it out from their website. Does GP-B have results that are either consistent with or not consistent with the geodetic effect or frame dragging or gravitomagnetism? Or is it that the unexpected errors swamped the difference in gyro alignment that would have been apparent if these GR effects were happening to the gyros.

    It's been months since April 2007 and years since they collected the data. Has GP-B produced results that confirm or falsify the predicted effects of GR? Or not? Or not yet?
     
  2. jcsd
  3. Dec 11, 2007 #2

    Garth

    User Avatar
    Science Advisor
    Gold Member

    You find this discussed here at length.

    Basically there were two unexpected experimental errors that swamped the results. The team have been modelling these errors using two different methods, geometrical and algebraical, in order to be able to subtract them from the data. Having two different methods means they can cross check the results.

    The first results published in April 2007 confirmed the GR geodesic precession prediction to within 1[itex]\frac{1}{2}[/itex]%, which was consistent with GR given the remaining error bars on the results. The much smaller frame-dragging result was then still swamped by noise.

    The full results are now expected in May 2008 (I think they must be paid by the day :biggrin:).

    Very interestingly in the Dec Mission Update there was a very intriguing diagram uploaded that showed the recent detailed results of 85 days worth of data and gave a result that was inconsistent with GR at the 1 [itex]\sigma[/itex] confidence level (68%).

    This has now been replaced by a similar diagram which makes no such claim!!

    Are they hiding something? :wink:

    We wait for the most accurate results next May with some anticipation!

    Garth
     
    Last edited: Dec 11, 2007
  4. Dec 11, 2007 #3

    Chris Hillman

    User Avatar
    Science Advisor

    See also [post=1303301]this old PF post[/post] by myself.
     
  5. Dec 12, 2007 #4

    rbj

    User Avatar

    i saw that back then, Chris. i was hoping for news of developments since. it seems to me that because of the unexpected errors, that GP-B might have failed to detect either the presense or absense of frame dragging. maybe they'll be able to rescue the subtle differences out of additive errors that are much bigger, but i'm starting to be less hopeful.
     
  6. Dec 12, 2007 #5

    Chris Hillman

    User Avatar
    Science Advisor

    Everyone must wait and see. Science is tricky and these experiments are looking for very small effects, so given the previously unmodeled effects which must be subtracted, it is perhaps not so surprising that analysis is taking so long.
     
  7. Dec 13, 2007 #6

    Garth

    User Avatar
    Science Advisor
    Gold Member

    At the April APS conference they were confident that they will be able to remove the unexpected noise in the data and obtain both precessions to an accuracy of [itex]\pm[/itex]1mas.

    Initially they had designed the experiment to an accuracy of [itex]\pm[/itex]0.1mas, so they have lost an OOM.

    Nevertheless that will confirm, or otherwise, the GR frame dragging prediction to within [itex]\pm[/itex]2% and the geodetic precession to within [itex]\pm[/itex]0.02%.

    Garth
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Can someone tell me if GP-B has results about GR, one way or the other?
  1. GP-B scoop (Replies: 1)

Loading...