vibration and shock

Go To Last Post
41 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi guys

I want to put a timer on a little rocket, just wondering if a AVR chip can handle like 50-100g (estimate) acceleration and the vibration?

I am also concerned about the xtal crystal, will it function correctly in term of accuracy of timing under these stress?

if not, any solution?

Thanks guys!

edit:

all these will happen under -20 degree C, thanks again!

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just go for it! I estimate that there is need to worry as long as you don't stand directly under the return to ground path! As the rocket is small I estimate that it will be low cost and expendable. Be sure to wear woolen under garments as I estimate that you will feel the cold.
Also be sure that you have some telemetry so that you can record flight progress & report on actual acceleration.

Report your experience here!

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'd be suspicious of the crystal handling that amount of acceleration. You might want to look at using a resonator instead of a crystal.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DS3231M is a RTC with a MEMS resonator (32kHz, 1Hz, 5ppm=0.4s/day).
Silicon Labs announced MEMS oscillators.
IDT also has some.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I'd be suspicious of the crystal handling that amount of acceleration. You might want to look at using a resonator instead of a crystal.
Or the internal RC oscillator perhaps? Calibrate it beforehand if accuracy is required.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Quote:

I'd be suspicious of the crystal handling that amount of acceleration. You might want to look at using a resonator instead of a crystal.
Or the internal RC oscillator perhaps? Calibrate it beforehand if accuracy is required.

I would prefer not to calibrate it, I prefer to use a accurate clock to start with.

Because the person using it is likely don't like to do that.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
I'd be suspicious of the crystal handling that amount of acceleration. You might want to look at using a resonator instead of a crystal.

Thanks I will look into it.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
DS3231M is a RTC with a MEMS resonator (32kHz, 1Hz, 5ppm=0.4s/day).
Silicon Labs announced MEMS oscillators.
IDT also has some.

Never use MEMS oscillators before, it looks good on specs, thanks.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Because the person using it is likely don't like to do that.

You misunderstand, the calibration I'm talking about is something you, the developer do on your work bench before you give this device to the rocketeer. You use a frequency generator or other accurate timing reference. This gives you an OSCCAL value that makes the AVR run at an accurate frequency. You write this to the EEPROM of the AVR. At next power on the AVR loads this value from EEPROM and writes it to OSCCAL.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Cliff, the main reason against using the RC oscillator is the temperature change due to altitude. It could be 25C at launch and -20C a couple of seconds later. This is going to affect an RC osc significantly.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ah good point (I should know this, I have a pilot's licence!)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The temperature depends on the profile and speed.
Can be enough temperature due to heat from the nosecone (atmospheric drag) and the motor.
If it spends enough time in the tropopause then the temperature could drop to -65C for a bit.
If it's suborbital then a lot colder than that.
PongSat

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

100Gs would be an acceleration of 3200 ft per sec per sec. .6 miles per sec per sec. 2181 miles per hr per sec. After 3 sec you are going 6000 mi per hr. I'm skeptical. Somewhat.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Ah good point (I should know this, I have a pilot's licence!)

Nice, I want to get a pilot license too!

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
100Gs would be an acceleration of 3200 ft per sec per sec. .6 miles per sec per sec. 2181 miles per hr per sec. After 3 sec you are going 6000 mi per hr. I'm skeptical. Somewhat.

Oops, it looks like I have done something wrong here, thanks for pointing it out.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Perhaps a definition of:

Quote:
on a little rocket

would be helpful.

Size, weight, number of stages, estimated peak altitude.

A photo would be Great.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Although Atmel does not give maximum acceleration specs, I would be amazed if there were any issues surrounding the processor itself at those accelerations. A through-hole device might be preferable, for peace of mind, or SMD that is also glued into place. Also, it would be helpful to know just how accurate things need to be. In most cases, timers on a rocket could be within the frequency spread of the internal RC without causing any particular problems. The burn time of the motors are not terribly accurate, atmospheric conditions will effect the launch, etc., perhaps there is a reason this will need higher accuracy however.

Also, the expected flight profile could be very helpful. The amount of accumulated error depends heavily upon both the amount of time the system runs as well as how far off the clock is. If either is a small number, there is nothing to worry about.

Martin Jay McKee

As with most things in engineering, the answer is an unabashed, "It depends."

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
100Gs would be an acceleration of 3200 ft per sec per sec. .6 miles per sec per sec. 2181 miles per hr per sec. After 3 sec you are going 6000 mi per hr. I'm skeptical. Somewhat.
Most model rockets with chemical engines fire for only a few seconds or less. Water-rockets fire for under a second. Accelerations of > 50G are not unusual. Vibrational and end-of-flight impact accelerations can easily exceed 100G.

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
Perhaps a definition of:
Quote:
on a little rocket

would be helpful.

Size, weight, number of stages, estimated peak altitude.

A photo would be Great.

JC

I don't have the files with me now, but I can report back when I get home.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

mckeemj wrote:
Although Atmel does not give maximum acceleration specs, I would be amazed if there were any issues surrounding the processor itself at those accelerations. A through-hole device might be preferable, for peace of mind, or SMD that is also glued into place. Also, it would be helpful to know just how accurate things need to be. In most cases, timers on a rocket could be within the frequency spread of the internal RC without causing any particular problems. The burn time of the motors are not terribly accurate, atmospheric conditions will effect the launch, etc., perhaps there is a reason this will need higher accuracy however.

Also, the expected flight profile could be very helpful. The amount of accumulated error depends heavily upon both the amount of time the system runs as well as how far off the clock is. If either is a small number, there is nothing to worry about.

Martin Jay McKee

I am not familiar with the mechanical stuff, as there is a mechanical guy dealing with it. My part is to make a timer and some other stuff.

    as of the timer, I want +/- 10ms minimum accuracy. the temperature will be under -20 degree C
    The launch will be on a mountain.
    At the moment will be one stage, but have plan to have a second stage if the timer thing is working as expected
.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

It could be 25C at launch and -20C a couple of seconds later. This is going to affect an RC osc significantly.

So, the temperature of the operating AVR, in all probability inside an enclosure along with other electronic components, would experience such a rapid temperature change? I think not. (That is not to say that the AVR might well be subjected to temperature extremes and that might affect internal oscillator accuracy.)

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
as of the timer, I want +/- 10ms minimum accuracy.
That's not a measure of accuracy, more a measure of precision.

Plus/minus 10 mS over what? Over a seconds? A minute? A day? The duration of the flight? Accuracy is often expressed in ppm. Crystals with ratings of +/- 50 ppm and better are readily available. That's +/- 3 mS per minute, or +/- about 4 seconds per day.

In contrast, with the internal oscillator user-calibrated to +/- 1% (10,000 ppm), you can expect +/- 600 mS per minute. This of course will shift with changing temperature, but so will a crystal oscillator.

JJ

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
Quote:

It could be 25C at launch and -20C a couple of seconds later. This is going to affect an RC osc significantly.

So, the temperature of the operating AVR, in all probability inside an enclosure along with other electronic components, would experience such a rapid temperature change? I think not. (That is not to say that the AVR might well be subjected to temperature extremes and that might affect internal oscillator accuracy.)
Back in the '90s, I was the secretary of the CAFE Foundation (if you are a pilot, you may have heard of it). One of the other board members designed and built a 68HC11 based barograph that was used for flight test data collecting (airspeed, altitude, temp, etc.). To keep the pressure transducers away from temp variations, they were enclosed in an insulated chamber that was kept within one degree of 100 degrees F.

It was surprising to notice how much more energy it took to keep them warm with 200 mph cold wind blowing past the airplane! :shock:

But I'll bet if the PCB is inside a block of foam, you can make the RC osc. work fine.... :wink:

Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away. Antoine de Saint-Exupery (1900 - 1944)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
as of the timer, I want +/- 10ms minimum accuracy.
A GPS or GLONASS receiver will produce a 1 pulse-per-second (PPS) signal with an accuracy of 1 microsecond.
Some time servers characterize this to an order of magnitude better.
An alternate is to use a receiver for a time signal radio station; this obtains a very accurate UTC time.
Sync the AVR to that pre-launch.
If the AVR is synced during rocket assembly then will have to estimate the worst drift until end-of-flight.
Try to have a rocket umbilical connection, or an umbilical-like connection, to get the PPS signal into the AVR.
bug13avr wrote:
the temperature will be under -20 degree C
The launch will be on a mountain.
Yep, gets that cold, and more, on a mountain in winter.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

joeymorin wrote:
bug13avr wrote:
as of the timer, I want +/- 10ms minimum accuracy.
That's not a measure of accuracy, more a measure of precision.

Plus/minus 10 mS over what? Over a seconds? A minute? A day? The duration of the flight? Accuracy is often expressed in ppm. Crystals with ratings of +/- 50 ppm and better are readily available. That's +/- 3 mS per minute, or +/- about 4 seconds per day.

In contrast, with the internal oscillator user-calibrated to +/- 1% (10,000 ppm), you can expect +/- 600 mS per minute. This of course will shift with changing temperature, but so will a crystal oscillator.

JJ

Hi JJ

I meant to say +/-10ms over approx. 2 minutes, I am hoping to find some sort of clock source with low temperature coefficient and reasonable tolerance, as absolute time is not important as long as it is stable over different condition, eg shock, vibration and different temperature.

Hope I am making sense, please feel free to point it out if I am not making sense :)

I would like to not use a RC osc. as the temperature is not always under -20 degree C, some days maybe warmer, and some days maybe colder.

MEMS resonator is interested, although I haven't look into it, I am just checking back between classes.

Zhuhua Wu - Electronic Engineering Student

Last Edited: Mon. Jul 29, 2013 - 01:52 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
bug13avr wrote:
as of the timer, I want +/- 10ms minimum accuracy.
A GPS or GLONASS receiver will produce a 1 pulse-per-second (PPS) signal with an accuracy of 1 microsecond.
Some time servers characterize this to an order of magnitude better.
An alternate is to use a receiver for a time signal radio station; this obtains a very accurate UTC time.
Sync the AVR to that pre-launch.
If the AVR is synced during rocket assembly then will have to estimate the worst drift until end-of-flight.
Try to have a rocket umbilical connection, or an umbilical-like connection, to get the PPS signal into the AVR.
bug13avr wrote:
the temperature will be under -20 degree C
The launch will be on a mountain.
Yep, gets that cold, and more, on a mountain in winter.

I have no experience on the technique you mentioned, but I am guessing GPS modules are expensive, and it will blow our budget, but thanks for the idea tho.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Go do some google searching about crystals and g forces etc. Many modern crystals can easily deal with 100G (or a lot higher shock loads). Orientation of the crystal so the G force is applied in the optimal direction will help with frequency errors etc.

Lots of info out there if you want to search, read and learn.

GPS receivers also have crystal oscillators - so using one of those won't necessarily resolve any mechanical failure worries...

cheers,
george.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
... but I am guessing GPS modules are expensive ...
With antenna these are about 40USD to 60USD.
A time signal radio receiver (module plus antenna) would be about 20USD to 30USD.
Atomix® Timepiece Kits, Plans & Movements
Build a WWVB Radio Controlled Nixie Clock

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

georges80 wrote:
GPS receivers also have crystal oscillators - so using one of those won't necessarily resolve any mechanical failure worries...
There are GPS receivers that can survive a rocket launch but GPS receivers are not necessary if the flight time duration is short.
Usually sync the rocket's time with an external time source then launch.
The external time source can be as simple as an NTP client from an NTP server.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
georges80 wrote:
GPS receivers also have crystal oscillators - so using one of those won't necessarily resolve any mechanical failure worries...
There are GPS receivers that can survive a rocket launch but GPS receivers are not necessary if the flight time duration is short.
Usually sync the rocket's time with an external time source then launch.
The external time source can be as simple as an NTP client from an NTP server.

Like I wrote... "won't necessarily..."

The KEY issue here is the survivability of the electronics which means that the OP needs to determine all the weak points in the design and address ALL of them. Adding a 'GPS module' does not solve his problem unless that 'GPS module' specifically can withstand the g-forces involved. In that case it would be a lot easier to just spec a crystal that can deal with it ALONG with all the other components.

Understanding how a crystal will deal with the g-forces (failure or frequency variation) will allow for a better implementation. At least with just a discrete crystal he can determine the effect more directly - with a GPS module he would need to understand how the crystal is oriented and request information that may/may not be available from the GPS module manufacturer.

The OP needs to DEFINE/SPECIFY what his timing accuracy and resolution needs to be before we tell him to use a GPS module or require him to sync his rocket clock to GPS time. In the case of GPS time (for accuracy) he also needs to know what the time offset is for reception of the time packet from the GPS engine (4800 baud etc etc). Again, just throwing a GPS module into the problem does not necessarily provide him with the 'ultimate' in timing accuracy without understanding all those fine details.

So, OP, what are the requirements (accuracy/resolution and actual calculated g-forces) that your timer needs to provide? Do you need absolute time accuracy or just resolution? How long is the entire timing interval?

Have you hit google and done some reading on crystal orientation and which crystals perform better at high g?

cheers,
george.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

georges80 wrote:

Lots of info out there if you want to search, read and learn.

Google is the first thing I do when I need answers, but maybe I am using the wrong keywords, I don't find much info on vibration and shock about crystal oscillator.

up to today, here are what I got:
http://www.greenrayindustries.com/library/AccSens_sglpage.pdf
http://www.ieee-uffc.org/frequency-control/learning/filler_paper.html

They are not very useful for me, I simply want something say I can handle blah blah vibration on a datasheet.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
DS3231M is a RTC with a MEMS resonator (32kHz, 1Hz, 5ppm=0.4s/day).
Silicon Labs announced MEMS oscillators.
IDT also has some.

Hi gchapman, thanks for you suggestion, I start looking into MEMS oscillators, it looks like they are going to be fit for my application.

ref: (go to slide two)
http://www.slideshare.net/SiTime/mems-oscillator-benefits-of-mems-oscillator-8680357

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
bug13avr wrote:
... but I am guessing GPS modules are expensive ...
With antenna these are about 40USD to 60USD.
A time signal radio receiver (module plus antenna) would be about 20USD to 30USD.
Atomix® Timepiece Kits, Plans & Movements
Build a WWVB Radio Controlled Nixie Clock

I am afraid I have to say $20 is too expensive, the rocket and whatever attach to it is designed to be expendable, the goal of the rocket is to deposite plants seeds to an area that is difficult/dangerous for human to access.

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
georges80 wrote:

Lots of info out there if you want to search, read and learn.

Google is the first thing I do when I need answers, but maybe I am using the wrong keywords, I don't find much info on vibration and shock about crystal oscillator.

up to today, here are what I got:
http://www.greenrayindustries.com/library/AccSens_sglpage.pdf
http://www.ieee-uffc.org/frequency-control/learning/filler_paper.html

They are not very useful for me, I simply want something say I can handle blah blah vibration on a datasheet.

Well, most datasheets for crystals have info on max vibration g forces in various planes and various frequencies. You aren't going to get a simple answer that says "good for a 100g rocket takeoff"...

cheers,
george.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

georges80 wrote:

Do you need absolute time accuracy or just resolution? How long is the entire timing interval?

I don't need absolute time accuracy, say I set my timer to be 1 unit of time, I want my timer NO.1, NO.2...NO.N, NO. (N+1) to be the same +/- 10ms, over about 2 minutes of fly time, in different temperature.

That 1 unit of time can be 100ms or 101ms, or 105ms.

So in all, I need a timer can't be killed/effected by the acceleration at launch, and a reasonably stable over a range of temperature.

Am I making sense?

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
Google is the first thing I do when I need answers, but maybe I am using the wrong keywords, I don't find much info on vibration and shock about crystal oscillator.
First search:First three hits:Some specs:
Applications
• GPS
• Agriculture
• Avionics
• Guidance
• Navigation
• Military
• Other

Features
• G-sensitivity down to 0.2ppb/G
• Low aging
• Up to 50,000G acceleration event survival
• Very good short term stability

JJ

"Experience is what enables you to recognise a mistake the second time you make it."

"Good judgement comes from experience.  Experience comes from bad judgement."

"Wisdom is always wont to arrive late, and to be a little approximate on first possession."

"When you hear hoofbeats, think horses, not unicorns."

"Fast.  Cheap.  Good.  Pick two."

"We see a lot of arses on handlebars around here." - [J Ekdahl]

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
I am afraid I have to say $20 is too expensive, the rocket and whatever attach to it is designed to be expendable, ...
The time signal receiver (a PPS generator) would be in the rocket's launcher (reusable).
If the rocket launcher is also expendable then I see your point.
Mountain avalanche abatement rocket launchers are either in fixed locations or are portable; these are reusable.
bug13avr wrote:
the goal of the rocket is to deposite plants seeds to an area that is difficult/dangerous for human to access.
Because a rocket has a speed or quantity advantage over a drone or helicopter?

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bug13avr wrote:
if not, any solution?
A possible alternate solution is Linear Technology's Silicon Oscillators (switched capacitors).
Pros:
1. Wide supply voltage range
2. Low supply current
3. Industrial temperature range
4. Low frequency drift versus temperature
5. Low ageing
6. Low price (2.09USD quantity 1)
Cons:
1. Accuracy - 10 times worse than the requirement.
2. Some drift versus voltage.

Accuracy can be compensated by measurement or calibration during assembly or pre-launch.
The drift due to supply voltage can be eliminated by using a regulated supply voltage.

LTC6930-X.XX - 32.768kHz to 8.192MHz Precision µPower Oscillators (Linear Technology)

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
bug13avr wrote:
I am afraid I have to say $20 is too expensive, the rocket and whatever attach to it is designed to be expendable, ...
The time signal receiver (a PPS generator) would be in the rocket's launcher (reusable).
If the rocket launcher is also expendable then I see your point.
Mountain avalanche abatement rocket launchers are either in fixed locations or are portable; these are reusable.
bug13avr wrote:
the goal of the rocket is to deposite plants seeds to an area that is difficult/dangerous for human to access.
Because a rocket has a speed or quantity advantage over a drone or helicopter?

Sorry I thought you mean mounting a $20 module on a rocket, now I get you.

Ha, why they are not using a drone or helicopter, I don't know,, that's not up to me, maybe they have a good reason to use a rocket, not maybe not. :)

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
bug13avr wrote:
if not, any solution?
A possible alternate solution is Linear Technology's Silicon Oscillators (switched capacitors).
Pros:
1. Wide supply voltage range
2. Low supply current
3. Industrial temperature range
4. Low frequency drift versus temperature
5. Low ageing
6. Low price (2.09USD quantity 1)
Cons:
1. Accuracy - 10 times worse than the requirement.
2. Some drift versus voltage.

Accuracy can be compensated by measurement or calibration during assembly or pre-launch.
The drift due to supply voltage can be eliminated by using a regulated supply voltage.

LTC6930-X.XX - 32.768kHz to 8.192MHz Precision µPower Oscillators (Linear Technology)

Thanks for your help, this one might be useful!

Zhuhua Wu - Electronic Engineering Student

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

joeymorin wrote:
bug13avr wrote:
Google is the first thing I do when I need answers, but maybe I am using the wrong keywords, I don't find much info on vibration and shock about crystal oscillator.
First search:First three hits:Some specs:
Applications
• GPS
• Agriculture
• Avionics
• Guidance
• Navigation
• Military
• Other

Features
• G-sensitivity down to 0.2ppb/G
• Low aging
• Up to 50,000G acceleration event survival
• Very good short term stability

JJ

Obviously your google skill is a lot better than me, thanks!

Zhuhua Wu - Electronic Engineering Student