Test of the Shipping Container Spartan IR Camera for the SOAR Telescope Miles P. Loh, Owen Y. Loh, & Edwin D. Loh Department of Physics & Astronomy Michigan State University, East Lansing, MI 48824 [email protected]

517 355–9200 x2480

29 December 2006 Abstract We report a test of the shipping container made by dropping it and a test mass having the same mass as the instrument. The shipping container contains bumpers that cushion the instrument. The natural frequency of the bumper and test mass is 5 Hz. An oscillation attenuates by a factor of 0.66 after one period. The maximum acceleratation amax depends on the height x of the drop as

amax = 4.3 g [( x − x0 )/(30 − 1.3) cm]1/2 , where x0 is the compression at equilibrium. The shipping container meets ASTM Assurance Level 1 (a drop from 12 in [30 cm]), which is the most conservative standard for packages of this mass. The instrument can safely withstand the 5-g acceleration for a drop from this height. We estimate the maximum acceleration for the two weakest parts, the Henein flexible bearings and the A-frames that hold the cryo-optical box, to be 80 g and 70 g, respectively.

1

Introduction

In 2004, we built a shipping container1 of plywood and wood. In the meantime, the international community adopted phytosanitary standards for shipping containers that forbad raw wood. We built a new shipping container that uses plywood only. Here we describe tests of the new shipping container (Figure 1). 1

Loh, O., & Loh, E., 2004, Shipping Container, Spartan IR Camera.

2

REQUIREMENTS

Figure 1: Left: Base of the shipping container with the test mass, a water-filled barrel. Right: Detail of the bumpers.

2

Requirements

For shipments of this size and mass, there are three standards. ASTM Assurance Level 2 (a drop from 9 in [23 cm]) is commonly used.2 ASTM Assurance Level 3 (a drop from 6 in [15 cm]) is the low end. ASTM Level 1 (a drop from 12 in [30 cm]) is the conservative standard. In the discussion below, we find that the instrument will survive an acceleration of 50 g. A-frames support the cryo-optical box. They are made of 1.5×0.75 in G-10. The natural frequency of the COB supported by the spring of the A-frame is 300 Hz. The Aframes will buckle at 70 g. Rotation stages are used for all of the mechanisms. There is a specification for the maximum operating load, but not for non-operating conditions. The maximum load is 100 N in the axial direction and 50 N in the radial direction. The greatest mass, 4.4 kg, is that of the f/12 camera mirror and arm. When in the shipping container, the load is nearly axial. Therefore the maximum acceleration with which the rotation 2

2004, Burgess, G., MSU Packaging Department, private communication.

2

2

REQUIREMENTS

stage still operates is 2.3 g. For the filter wheels, the load will be radial when the instrument is shipped, but the mass is only 0.6 kg. We do not know the maximum allowed acceleration during shipping. Accidentally we dropped the f/21 collimator assembly 50 mm onto a granite table, and the rotation stage survived. We estimate that the acceleration at the rotation stage was 430 g.3 4-eye detector assembly has several features requiring thought. The Henein flexible bearings appear to be fragile. The load is 0.3 kg. The maximum acceleration is 290 g in the axial direction, and 60 g in the radial direction for 1000 cycles. Since the acceleration will be 45◦ from radial during shipping, the maximum acceleration is 80 g. The detector mounts must hold the detectors. A force of 40 N applied to one corner will pull the detector out of the mount, and the other three corners are secure. This is an acceleration of 200 g on the center for a mass of 90 gm. In the direction of the shipping container dropping, which is 80◦ to the normal of the detector, the maximum allowable acceleration is 1100 g. The detector printed circuit board has a cut that separates the cold and warm sections. The cold section is securely supported by the detector, but the warm section must support one end of a heat strap. The motion of concern is a twist of the bridge between the cold and warm sections. The board does not break with a force of 3 N applied at the heat strap. This translates to an acceleration of 45 g for a mass of 7 gm, half the mass of the heat strap. The natural frequency is 40 Hz. Furthermore, since the direction of acceleration during shipping is 80◦ to the normal of the board, the maximum acceleration is at least 270 g. Quartz strings hold each eye against the stop. Since the load is balanced with a counterweight, no extra tension is needed when acceleration is increased. Bolted joints can hold a large static force. For example, the three 6–32 bolts holding the f/12 collimating mirror can support 330 g (8000 N for a 2.6-kg mirror). The bolts on the f/21 collimator assembly were tested (accidentally) for shock. We dropped the assembly 50 mm onto a granite table, and the mirror arm hit the table. The shim moved 0.17 mm. For this very severe shock, the movement would not preserve optical alignment for most of the optics. We estimate that the shock on the mirror arm was 430 g. (See footnote.) 3

We used this model of the fall: The end of the mirror arm that supports the counterweight hit the granite table. That end is a beam, 3-mm wide, 24-mm long and 19-mm deep. The beam bends to stop the fall.

3

3

3

MEASUREMENT OF THE ACCELERATION

Measurement of the acceleration

To measure the acceleration, we lift the shipping container and a 210-kg test mass (a 55-gallon barrel filled with water) with a crane. Using a quick-release trigger (Figure2), we drop the apparatus suddenly. A CXL251Z accelerometer (Crossbow Technology, San Jose, CA) is mounted on top of the barrel. The range of the accelerometer is ±25 g, and its bandwidth is 0–100 Hz (–3 dB). • Drop from a height of 3 cm. See Figure 3. The system initially oscillates at 60 Hz for 0.2 s. (The frequency is clearer on the spectrum rather than on the time series.) Over a longer period, the system oscillates at 5 Hz. The oscillation at 5 Hz is the vertical motion of the mass, a water-filled barrel, on the spring, the bumpers. We show later that the oscillation at 60 Hz is most likely due to the barrel and water. • Drop from a height of 10 cm. Let us examine the time series (Figure 4). The system is dropped at –.16 s. It is in free fall until it hits the ground at –0.03 s. (In free fall, Figure 2: Quick-release trigthe drop is 1.3 cm less than the height of the drop, since the ger. A person holds the other bumpers extend without a load.) Between 0.08 and 0.23 s end (far to the right, not in the picture) of the wooden and between 0.38 and 0.44 s, the system is again in free beam with a rope, which may fall; it leaves the ground. be released quickly. BeThere is a brief, sharp acceleration at 0.00 s. The accelcause the straps that hold the eration is clearly sharper than the 60-Hz oscillations; in the weight are very close to the lower-left panel of Figure 4, compare this event with the sur- hoist, the person supports litrounding ocsillations. The bolts that hold the bumpers lie on tle weight. the center line of the bumpers. The bolts at the upper and lower parts of a bumper hit before the bumper bottoms out. The sharp spike is caused by the bolts hitting. This event occurs near the maximum of the acceleration filter with a low-pass filter; it occurs when the bumpers are most compressed. The spectrum shows a peak at 5 Hz and at 60 Hz. The 60-Hz oscillation is not due to the interaction between the shipping container and the load; it is due to the load itself. The 60-Hz oscillation occurs both in free fall and when the apparatus is on the ground. If it were caused by interaction between the shipping container and the load, its frequency would change by a factor of 3.7 when the mass changes from that of the base (15 kg) to that of the instrument (210 kg).

4

3

MEASUREMENT OF THE ACCELERATION

2

2

1.5

1.5 1 a
g

a
g

1 0.5 0

0
0.5


0.5


1


1
1

0.5

0

1 t
s

2


1.5
0.2
0.1

3

0

0.1 0.2 0.3 0.4 0.5 t
s

5

10 20 f
Hz

10

0.4

5

a
gHz12 

a
g

20 0.6

0.2 0
0.2
0.4

2 1 0.5 0.2 0.1


1

0

1 t
s

2

1

3

2

50 100 200

Figure 3: Time series and spectrum (bottom right panel) of the acceleration for a 3-cm drop. In the lower left panel, a low-pass, 10-Hz filter is applied.

5

MEASUREMENT OF THE ACCELERATION

4

2

2

1.5 a
g

a
g

3

0

1 0.5


2

0
0.5


4
1

0

1 t
s

2


1
1

3

0

1 t
s

2

3

4 a
g

2 0
2
4
0.2

0

0.2

0.4 t
s 20

4

10

a
gHz12 

6

a
g

2 0
2
4
6
0.1

0.6

0.8

1

5 2 1 0.5


0.05

0 t
s

0.05

0.1

1

2

5

10 20 f
Hz

50 100 200

Figure 4: Time series and spectrum (bottom right panel) of the acceleration for a 10-cm drop. In the upper right panel, a low-pass, 10-Hz filter is applied. The system is dropped at –.16 s. It is in free fall until it hits the ground at –0.03 s. Between 0.08 and 0.23 s and between 0.38 and 0.44 s, the system is again in free fall. There is a brief, sharp acceleration at 0.00 s.

6

3.1

Conclusions

3

MEASUREMENT OF THE ACCELERATION

• Drop from a height of 35 cm. See Figure 5. Since the average acceleration is –1 g during the interval –0.26 to 0 s, the system falls freely during this time. From 0 to 0.04 s, several sharp positive accelerations occur; these are due to the bumper bolts hitting each other. From 0.08 to 0.44 s, the average acceleration is again –1 g; the apparatus bounced up. The next collision with the floor occurs at 0.44 s. The peak in the spectrum (Figure 5) at 2.2 Hz is due to the first collision with the floor, a bounce, and the second collision with the floor. The other bounces cause peaks between 2.2 and the fundamental bumper-mass oscillation at 5 Hz. In addition there are harmonics of the multiple collisions.

3.1

Conclusions

The sharp spikes appear at an acceleration of 4 g. There is one for the drop of 10 cm. For the drop of 35 cm, eight occur on the first collision; 7, on the second; and one, on the third. These are probably caused by the bumper bolts hitting each other. There are eight possible occurances, one for each bumper. For the drop from 35 cm, bottoming out causes the acceleration to increase at least by a factor of 2. The factor may be larger since the sensor attenuates high-frequency acceleration. We can ignore the oscillation at 60 Hz, since the instrument feels only the acceleration between the base of the shipping container and the instrument. We are able to meet ASTM Assurance Level 3 (a drop from 6 in [15 cm]), which is the low end. The plan is to meet Assurance Level 1 by redesigning the mounts for the bumpers.

7

3.1

Conclusions

3

MEASUREMENT OF THE ACCELERATION

6

50

5 a
gHz12 

a
g

4 3 2 1 0

10 5 2


1
1

a
g

20

0

1 t
s

2

1

3

1

5

10

50 100

500

f
Hz

25 20 15 10 5 0
5
0.3


0.2


0.1

0

0.1

0.2

a
g

t
s 25 20 15 10 5 0
5
0.05

0

0.05 t
s

0.1

0.15

Figure 5: Time series and spectrum (top right panel) of the acceleration for a 35-cm drop. For the top left panel, a 10-Hz low-pass filter is applied to the time series. The middle panel shows two periods where the apparatus is in free fall, where the average acceleration is –1 g. During the interval –0.26 to 0 s, the system is yet to hit the floor. The system bounced up off the floor between 0.08 to 0.44 s. The bottom panel shows the region where the apparatus hits the floor for the first time, which occurs between 0 and 0.03 s. Several sharp positive accelerations occur; these may be due to the bumper bolts hitting each other because the bumpers bottomed out. The peak acceleration is close to the saturation of the accelerometer.

8

4

4

ACCELEROMETER LOCATION

Accelerometer location

a
gHz12 

a
gHz12 

Where should the accelerometer be mounted? Mounted on the top of the bar25 rel, the accelerometer is not completely 20 coupled to the test mass. The sides of 15 the barrel may move up and down without moving the entire mass of the water. 10 Mounted on the plywood base, the ac5 celerometer measures the acceleration of the base and less directly the acceleration 0 0 50 100 150 200 of the mass of the test mass. The base f
Hz can deform and move the test mass by a 25 much smaller amount, since the mass of the base is much smaller. 20 We made several measurements with 15 the accelerometer mounted to the ply10 wood base near the bottom of the barrel. The acceleration is larger in the frequency 5 range 10–80 Hz (Figure6), whereas the 0 oscillation of the test mass is at 5 Hz. 0 50 100 150 200 Because of this argument and these f
Hz measurements, we decided to put the acFigure 6: Spectrum of the acceleration for a 30celerometer on the top of the barrel. cm drop with the accelerometer mounted on the base (bottom panel) and on the top (top panel).A filter is applied to reduce the point-to-point noise.

9

5

5

NEW BUMPER MOUNT

New bumper mount

For the old bumber mounts, the bolts holding the bumpers hit before the bumpers bottom out; that problem is fixed with the new bumper mounts by offsetting the bolts (Figure7).

Figure 7: New bumper mount. Note that there is no central bolt where two bumpers join and that there the bolts for the top bumper are offset from those of the bottom bumper.

The test at Assurance Level 1, a drop from a height of 30 cm, shows that the sharp acceleration, present with the old bumper mount, does not occur. Figure 8 shows the full time series of the acceleration. The bottom panel of Figure 9 shows the acceleration at the time when the apparatus first hits the floor. There are no accelerations that are unresolved in time as there were with the old bumbers mounts (Figure 5). The high-frequency accelerations are not caused by interaction between the mass and the floor. They are present when the apparatus is falling freely. See the highfrequency accelerations in the middle panel of Figure 9, which shows the acceleration at the time when the apparatus is first released and falling freely. When the apparatus starts free fall or hits the bumbers, the high-frequency oscillations occur for a period of about 0.1 s.

10

5

NEW BUMPER MOUNT

15

a
g

10 5 0
5
0.3


0.2


0.1

0 t
s

0.1

0.2

0.3

0.4

0.5

0.6 t
s

0.7

0.8

0.9

15

a
g

10 5 0
5 0.3 3 a
g

2 1 0
1
2 1

1.5

2 t
s

2.5

3

Figure 8: Time series (filtered with a 200-Hz, low-pass filter) of the acceleration for a 30-cm drop with improved bumper mounts. Between –0.16 and –0.06 s, the release mechanism limits the acceleration to -0.2 g. The apparatus is in free fall in the interval (–0.06, 0.16) s, since the average acceleration is –1 g). The bumpers engage in the interval (0.16, 0.28) s. The apparatus bounces up in the intervals (0.28, 0.62), (0.74, 0.96), (1.12, 1.24), and (1.4, 1.47) s.

11

5

50

4 a
gHz12 

3 a
g

NEW BUMPER MOUNT

2 1 0

20 10 5 2


1
1

0

1 t
s

2

1

3

1

5

10

50 100

500

f
Hz

2

a
g

0
2
4
6
0.1


0.05

0 t
s

0.05

0.1

0.1

0.15

0.2 t
s

0.25

0.3

a
g

15 10 5 0

Figure 9: Time series with a 10-Hz low-pass filter applied (top left) and spectrum (top right) of the acceleration for a 30-cm drop with improved bumper mounts. The expanded time series (with no filter) shows the time after release (middle) and at the first collision with the floor (bottom).

12

6

6

CONCLUSION

Conclusion

The natural frequency of the bumper and mass is 5 Hz. An oscillation attenuates by a factor of 0.66 after one period. The test mass and base have complex oscillations in the range 35–200 Hz. We ignore these because they are present even in free fall. Because the instrument is stiffer than the barrel of water, we expect these oscillations not to occur when the instrument is mounted on the shipping container. The maximum acceleratation amax depends on the height x of the drop as

amax = 4.3 g [( x − x0 )/(30 − 1.3) cm]1/2 , where x0 is the compression at equilibrium. (See Figure 10.) Data with the accelerometer on the base and with bolts hitting were not used in the fit.

7 S

6 B amax
g

5 4

S

B

3 2

B

1 0 0

5

10

15 20 25 height
cm

30

35

Figure 10: Maximum low-frequency acceleration for a drop from height with original (point) and improved bumper mounts (cross). The acceleration was measured with a 1-pole, 10-Hz low-pass filter. An “S” indicates a case where the bumper bolts bottomed out. A “B” indicates cases where the accelerometer in mounted on the base, rather than the top of the barrel. The line is a best fit for the data with the accelerometer on the top and the bolts not bottomed out.

The shipping container meets ASTM Assurance Level 1 (a drop from 12 in [30 cm]), which is the most conservative standard.

13