In September 2005 (for the 50th anniversary of the atomic clock and 100th anniversary of the theory of relativity) we took several cesium clocks on a road trip to Mt Rainier; a family science experiment unlike anything you've seen before.
By keeping the clocks at altitude for a weekend we were able to detect and measure the effects of relativistic time dilation compared to atomic clocks we left at home. The amazing thing is that the experiment worked! The predicted and measured effect was just over 20 nanoseconds.
The main Project GREAT page has technical details, but below are some photos and commentary on the trip.
Dad makes final connections to the clocks. It took about half an hour to bring all the equipment down from the upstairs lab out to the car.
The kids took the back seat.
Here's a view of the car. What a mess. The three clocks fit on a middle seat in the van.
The front seat area is where I put the counters, laptop, inverters, power monitors, environmental sensors.
It comes to 400 pounds of batteries; 200 pounds of clocks.
A view of the middle seat.
Not very elegant but we were in a hurry.
The nylon straps were probably a good idea.
The seats themselves provided excellent vibration isolation.
The power strip provides 120V AC from DC inverters to the clocks. The clocks also had 24 VDC backup power, as well as internal batteries in each cesium clock.
This NASA-like triple redundancy is to make sure the clocks don't stop.
eight 70 Ah 12 V marine batteries arranged in four pairs of
two. Two pairs provide 24 VDC, one pair 12 VDC, and one spare pair.
I called them bus A and bus B. I monitored both of them and secretly hoped to also experience a "Main B bus undervolt" (as in Apollo 13). But, no, there were no AC or DC power problems.
The 12 VDC bus is used for the two AC inverters. Next time I will use pure sine wave inverters. The 5071A sing with modified sine waves.
|View of front seat showing the three HP 53132A
TIC (time interval counters), Vaisala temp/humidity meter, laptop, and lots
of wires. Each TIC has a AC power cord, a DB9 RS-232 cable, and two coax
You can also see the red/black 12 VDC power lines coming from the engine (alternator) and going to the back seat (battery bank)
|Another view of the three clocks. There was no time to place them picture perfect in the car seat. I did place hard foam between the clocks so as not to scratch them.|
|To keep everything straight I color coded the
clocks, cables, and counters: red, green, and blue.
As a special treat the kids got Jones Sodas at the beginning of the trip; you guess it - fufu berry red, apple green, and bubble gum blue.
Those of you with more than one child will understand why the success of a long road trip depends on equality. Three kids, three clocks, three sodas.
|Mom says goodbye to us. She stayed behind to
study for her nursing board exams.
We would come back about 20 ns older compared to her.
Or, the other way to look at it (since this is relativity after all), is that she would become 20 ns younger than us upon our return. Note to husbands: this could be a useful gift idea for your wife.
|This view from the driver's seat allowed me to check on the status of the experiment while I was, er, um, driving.|
|I had 10 command windows open during the trip.
COMLOG is a specialize time-stamped serial port logging tool that I use.
Three ports were used to periodically log the 5071A status page (the SYST:PRINT? command)
Three ports were used to capture the 1 Hz pair-wise phase readings of the clocks.
One port was used to log the HMP sensor.
One port was used to log the Garmin NMEA GPS data stream.
Keeping an eye on the clocks as we drove.
It turns out there were no issues with vibration or shifting.
|The kids have seen Super-Size Me so we
don't stop at McDonald's anymore.
There wasn't a big hurry to get up to Mt Rainier; we were on vacation. But we got a late start and every hour spent at low altitude meant one less hour spent at high altitude.
I figured the clocks would see about 1/2 nanosecond of time dilation for every hour up on the mountain.
So this 10 minute stop at DQ cost me $8 and 83 picoseconds.
|For those of you with a map; on the way down we
took the long way using highways 507 and 702.
Days later I realized there was a shorter route to the mountain. But in all the excitement to get the mobile time lab rolling I forgot to check the map for optimum route.
|Happy Campers, 01 hour, 39 minutes into the UTC day.|
As evening began, we filled up at the last gas station before the final road into the park and the twisting road up the mountain.
By this point in the trip I was considering keeping all the clocks running in the van all weekend instead of carrying yet again all the heavy gear up to the hotel room. So I figured filling up was prudent.
|It was getting dark by the time we got to the park.|
|Top of the road, at about 5400 feet.|
|Nice early morning colors on the upper mountain.|
|A full moon during the weekend.
The moon has a 28 day cycle (27.32166155 to be more exact). That comes out to 4.236e-7 Hz.
A time to reflect on the contrast between using a 28-day lunar cycle and a 9 192 631 770 Hz Cesium resonance to keep track of time.
|By contrast inside this van are a bunch of
cesium atoms resonating at 9192 MHz
That's 9192 MHz
Compare that to the moon's 423 nHz.
|Fun 'n games.
We had a lot of time to kill while waiting for the clocks to do their thing.
Waiting as cesium clocks undergo time dilation is much worse than watching a kettle boil.
|The Lodge sure knows how to do breakfast.|
|Beth climbs Mt Rainier.|
|A view of Paradise Lodge, Mt Rainier, elevation
5400 feet (1640 m).
You can see our green van parked in front.
|Nice couch for reading.
While the kids counted pages I counted picoseconds.
A visit to the ranger station.
You know, the whole experiment could be done in reverse. By measuring the time dilation in nanoseconds one can later determine elevation by solving for h.
|Signs like this helped explain to people why a
car was locked and idling in front of the Lodge for two days.
For an hour once I forgot to lock the car. I came out later to find it locked and not running. Turns out a hotel employee who couldn't read turned the car off and took the keys to the front desk.
All the backup systems worked and the clock didn't miss a beat.
|Nice day for a hike.|
|And a rest.|
|The high point of the trip; Glacier Vista.|
|Oops. Forgot to check the gas gauge. Although
the car was just idling, it went through a tank of gas in less than two
I calculated it would run out by 8 AM Monday morning.
So I woke up at 430 AM and drove back down the mountain an hour to the nearest open gas station.
So 2+ hours later the car was back on the Lodge.
That cost me $50 in gas and 1 nanosecond of time.
|Pre-dawn trip to fill up on gas.|
|A nice sunrise photo of Mt Rainier.|
Leaving the park entrance, near Ashford, WA.
We stayed until I calculated we had a full 40 hours of exposure at mile-high elevation. I added a few extra hours to make up for lost time due to the unexpected gas trip.
|On the way home.
48 hours later, we were all tired and sick of the experiment.
Checking on the clocks every hour or so, day and night, I was most tired of all.
The only thing that kept me going was knowing that the experiment hadn't screwed up yet, and that within minutes of returning home we'd know if we saw time dilation or not.
|Back home again.|
Amazingly, we did see time dilation.
Here's a final graph with 8 days of data. This represents 3 days pre-trip, the 2 day Mt Rainier trip, and 3 days post-trip.
You can see the blue clock was a little weird. Not sure what caused that yet.
But the time dilation was somewhere in the 20 to 30 ns range. The number we expected was 23 ns so I'm very pleased with the result.
Send comments or questions to: tvb
Return to LeapSecond.com home page.
Project GREAT: General Relativity Einstein/Essen Anniversary Test