Hubble Presentations
Meteoroid and Debris Impacts on the WF/PC I Radiator
Donald H. Humes / NASA Langley Research Center
William H. Kinard / NASA Langley Research Center


We spent two days in the clean room at Goddard examining the WF/PC-I radiator with a microscope to measure the damage done by meteoroids and man-made orbital debris during its 3.6 years in orbit.

It was a difficult job - moving the microscopes around on a heavy stand, positioning it close to the WF/PC I radiator while never touching it, and looking through the microscopes while in awkward positions on the steps of a ladder.



WF/PC-I radiator in clean room at NASA GSFC, with microscope in front.

I (Don Humes) looked at every impact site examined, but Mark Kulick, a Lockheed meteoroid and debris researcher, made all the measurements. Only a few photographs were taken because of the difficulty in taking them in the short time we had.

We were aided immensely by the survey done at Goddard using theodolites to obtain the location of possible impact sites. Goddard provided the coordinates of 100 possible impact sites, assigned a number to each, and rated them by size on an arbitrary scale of 1 to 10 (10 being the largest).

Outline



I will talk about (1) the crater flux based on the number of craters of various sizes found in the aluminum radiator plate, (2) the spallation of the ZOT paint in an area around the craters, (3) rings seen in paint on the LDEF but not seen in the ZOT paint and (4) the shape of the craters in the aluminum.

Comparison of the damage seen on the WF/PC I radiator to damage seen on the LDEF will be made throughout the talk.

Crater Flux

The theodolite survey data was given to me by Henry Sampler and also had the name of Jerry Gay on it. They did a great job - finding impact sites with craters as small as 270 microns, that is .011 inches.

We examined 72 of the 100 possible impact sites and found 53 to be true impacts and 18 to be deposits of gooey particles or to be scrapes in the paint. At one site we found nothing. We examined all the size 4 to size 10 impact sites, and half of the size 1 to size 3 sites.

Radiator Surface

The WF/PC-I radiator was an aluminum plate, 0.160 inch thick, that was painted with ZOT paint. ZOT paint is a ceramic thermal control paint that gives a <0.18 and e>0.85 in the 0.28 to 2.50 micron range. The pigment is zinc orthotitanate and the binder is potassium silicate. The specs call for the ZOT to be 3 - 6 mils thick. A pre-flight inspection report lists the thickness as 6.8 mils. We measured a thickness of 4.3 mils at one place.

Large Impact Craters

The large craters in the aluminum were slightly to moderately irregular in shape and had pits (or sub-craters) of different depths inside the primary cavity.

The lips were not well formed like those in unpainted aluminum - having broken off or never developed.

There was a dark area around the crater where the raised lips would have been.

Condensed molten droplets, apparently of aluminum, were common near the top of the craters and at all depths.

The ZOT paint spalled off the aluminum plate in a large area around the craters.

The spall area was irregular in shape too.

The 14 impact sites that had craters with a diameter greater than 450 microns (measured at the aluminum plate surface) were of the type classified here as large craters. They had a single crater that, while not usually round, was not extended in any direction.



Crater on WF/PC-I radiator. Dimensions: lip, 500 x 540 microns; at plate surface, 370 x 460 microns; depth, 185 microns. Paint spall area around crater, 1320 microns. Cracks in paint can also be seen.

The very largest craters had nearly round rims, but with irregular bottoms.

The largest crater found in the WF/PC I radiator was 900 microns in diameter. It is nearly round and resembles craters found in unpainted aluminum on the LDEF, except that the lips are poorly developed.



Largest crater on WF/PC-I radiator. Dimensions: lip, 980 micron diameter; at plate surface, 900 micron diameter; depth, 360 microns. Paint spall area around crater, 5400 micron diameter (not shown)

Small Impact Damage Sites

Craters smaller than about 450 microns (there is no clear cutoff size) were highly irregular in shape with a number of cavities that were sometimes connected and sometimes not.

It appeared as if the ZOT paint acted like a meteoroid bumper, shattering the impacting particle before it hit the aluminum plate, and allowing the fragments to disperse somewhat. The surprising thing is that the fragments could disperse so much in such a short distance - the thickness of the paint.

Perhaps this shows that many meteoroids are a loose aggregate, a porous and fragile assemblage of grains, as Brownlee has suggested in a recent study of the densities of captured stratospheric meteoroids.

And I do suspect that most of the impacts on the WF/PC I radiator were caused by meteoroids and not by man-made debris, as I will discuss later.

Crater fields like this were seen on the LDEF plates also, but only very rarely (less than 1 percent of the impacts). Those crater fields were usually linear and showed evidence of the impact direction and seemed to be caused by highly oblique impact angles (greater than 80 degrees from the normal). Those impacts also suggested that many meteoroids are loose aggregates. The ZOT paint seemed to enhance the breakup and dispersion. The painted aluminum plates on the LDEF did not exhibit the type of cratering seen on the WF/PC-I radiator.

But there are other possibilities.

Secondary particles created when meteoroids and debris were fragmented while penetrating the solar panels could have struck the radiator. That would explain the dispersion of the fragments. But I think it is unlikely that the ejecta would have created small clusters of craters so widely separated from each other.

There was a highly irregular spall area around the crater field.

Measurements

The measurements we made were (1) the diameter of the crater at the aluminum plate surface, (2) the diameter at the top of the raised lips, when they existed in any form, (3) the depth from the plate surface to the deepest point and (4) the diameter of the spall area around the crater.

Because the large craters and their spall areas were irregular in shape, the diameters measured are crude measurements but did not require much judgement in assigning a representative diameter.

On the other hand, the measurement of the crater diameter at the small impact sites did require a judgement.

The depth was straightforward, and the spall diameter was straight-forward, but the crater diameter was not.

We chose to imagine a circle surrounding most of the crater field and call that the crater diameter. This may overestimate the size of the crater that would have produced in an unpainted plate - but that is what we measured.

In some cases the crater field was extended in one direction and the width and length of the crater field was measured.

There is a correlation between the Goddard size estimate and our measured crater diameter, although it is not perfect. A few entries are out of place.

We can expect that all the large craters were found and measured and that, in fact, essentially all the craters in the aluminum were found, although some were not examined.

Impacts that only damaged the paint and did not produce spallation would not have been found during the Goddard theodolite survey and hence were not examined by us.

Many size 1,2, and 3 possible impact sites were not examined. We can estimate from the statistics of those that were examined how many probably are impact sites and what size they probably are. Having done that we can estimate the flux of various size craters in the radiator.



Here then is the cumulative crater flux as a function of crater diameter, i.e. the number of craters larger than some limiting size per unit area per unit time.

The previous discussion about the Goddard size estimate and the correlation to our measured crater diameters was included to show that we have not made the best flux measurements that we could have made. We guessed at what size craters we would have found at some sites because we did not have time to look at them.

The production of small craters in the aluminum, those with a diameter less than the paint thickness, was probably inhibited by the presence of the paint.

But we expect that the diameters of the largest craters were not affected much by the paint and that the flux measurements for the three largest threshold crater sizes are close to what the crater flux in an unpainted aluminum plate would have been. We have evidence to support that from a painted aluminum plate from the LDEF.

LDEF EPDS Thermal Cover - Row 12

Crater Flux

The painted plate was on Row 12 of the LDEF. The cumulative flux of various threshold crater sizes on the that plate had 90 percent confidence limits.

The crater flux in the painted aluminum on Row 12 matched the crater flux in the bare aluminum on Row 12 for craters with a diameter greater than the paint thickness, and did not for a crater diameter less than the paint thickness.

In other papers, it is argued that most of these craters were caused by meteoroids, greater than 80 percent. That argument is based on the distribution of craters around the LDEF, on the chemical analyses of residue found in LDEF craters, and on measurements on the particulate environments in the 1960s when man-made debris presumably was not abundant.

And since the craters in the WF/PC I radiator are in the same size range and in the same flux range, we assume they also were caused mostly by meteoroids.

The paint on the LDEF was not the same as that on the WF/PC-I and I'll talk more about that later.

Crater Flux on Radiator

The fluxes were measured in unpainted aluminum on various sides of the LDEF. The HST was at the same altitude and inclination as the LDEF so the meteoroid fluxes should be directly comparable for the three largest crater flux measurements.

The problem is - we don't know what the orientation of the WF/PC I radiator was. If we knew its orientation history we could see if the WF/PC I radiator experienced the same flux as the LDEF.

If it turns out that the WF/PC I radiator was effectively randomly oriented or slightly more protected than a randomly oriented surface (protected by its orientation or by the solar panels) - then the crater fluxes on the radiator would be in agreement with the LDEF data, and then we could conclude that the LDEF and that secondary particles generated when the solar panels were penetrated were not significant. But that all hinges on knowing the orientation history of the HST.

Spallation of Paint

I would like to compare the spallation areas in the paint seen around the craters on the WF/PC I radiator with paint spallation areas seen on a plate from the LDEF.

The EPDS thermal cover on Row 12 of the LDEF was an aluminum plate painted with Chemglaze A276. The specs called for a thickness of 4.5 mils. We measured 2 - 4 mils at various locations.

Spallation of the paint on this LDEF plate usually occurred in one of three ways with a spall diameter two to nine times the crater diameter.

About sixty percent had a near circular spall area with no radial cracks outside the spall zone.

About nineteen percent had an inner spall diameter and incomplete spallation for a larger spall diameter - but with no cracks seen where spallation was incomplete.

About eight percent had a small spall diameter with both radial and circumferential cracks where an outer spallation zone was nearly formed.

Of the remainder of the spallation sites, less than four percent had irregularly shaped craters (none like those seen on the WF/PC-I radiator) and irregular spall areas. About six percent had paint in the crater. About two percent were dings - chipped off paint with no crater.

A typical irregular crater and spall area accounted for less than four percent of the impacts of the LDEF EPDS thermal cover. These craters were single, oblong cavities and did not resemble the multiple-cavity craters seen on the WF/PC-I radiator.

A spall on the WF/PC-I radiator look a lot like one of the LDEF spalls previously shown - a small spall area and radial and circumferential cracking outside the spall area. The WF/PC-I radiator spall area is not as round as that usually seen in the A276 paint on the LDEF.

On the LDEF plate, the diameter of the spall area, in proportion to the crater diameter, varied with crater size. The ratio of spall diameter to crater diameter decreasing with increasing crater size.

For the ZOT paint on the WF/PC I radiator the diameter of the spall area was also two to ten times the crater diameter.

But the spall area for large craters was much larger for the ZOT paint than for the A276 on the LDEF.

Also, the proportionate size of the spall area increased with increasing crater size for the ZOT paint - just the opposite of the trend for A276.

Ring of Chemglaze 9924 Primer Around the Crater

On the LDEF, there was a ring of the red Chemglaze 9924 primer around the crater where the raised lips would have been on an unpainted aluminum plate. While the two coats of white A276 did spall off the primer coat did not. Every where else in the spall area the primer coat spalled off the plate. The large craters on the WF/PC-I radiator also had dark rings around them where the raised lips would have been, but we could not tell if this was the primer paint.

A feature seen in the A276 paint on the LDEF but not seen in the ZOT paint on the WF/PC I radiator was a series of perfectly circular concentric rings around the crater, far outside the spall zone - perhaps 15 to 30 times the diameter of the crater.

The rings do not show up under normal room lighting. They only show up when lit from the side at an extreme angle and the room lighting is not too bright.

We looked for rings in the ZOT paint and did not find any, but we could not produce the lighting that would show them up best while we were in the clean room. We only had a flashlight to use for the extreme angle lighting.

However, we do not expect that such rings exist in the ZOT paint. The theory on the rings in the A276 is that atomic oxygen attacked the organic binder in the paint and it was removed, leaving the powedered white pigment on the surface. Shock waves created during an impact then reflected off the two surfaces of the aluminum plate and interferred, sometimes contructively, sometomes destructively, and caused the pigment to be thrown off at particular distances from the crater. The surface certainly has a white powder on it. You can see where it was wiped off near the edges of the plate during handling.

The ZOT paint may not be affected by atomic oxygen this way.

Crater Shape

Craters in unpainted aluminum are nearly hemispherical with lips that rise above the plate surface. Large craters are very nearly hemispherical, while small craters are slightly deeper than hemispheres.

Small craters in unpainted aluminum have a wide range of shapes that have a mean P/D of about 0.6, but as the craters get bigger they have little variation from hemispheres - at least at this location, which was Row 10 on the LDEF, near the RAM direction.

Some sides of the LDEF, like the space end, did not show this strong funnel shaped data field.

The craters in the painted aluminum on the LDEF were shallower than those in unpainted aluminum, as you would expect because of the thickness of the paint penetrated.

The small craters in the aluminum on the WF/PC-I radiator were very shallow compared to the diameter. That is related to the way the "diameter" was defined for the collection of cavities that were typical of the small impact sites on the radiator. This further illustrates that the small craters on the WF/PC-I radiator were very different from those on the painted aluminum plate from the LDEF.

Summarizing Remarks



HST WF/PC-I

Suggested Future Work on Radiator

Complete examination of possible impact sites identified by GSFC.

Look for impact craters in ZOT paint.

Obtain good photographs of impact damage.

Measure thickness of ZOT paint at many locations.

Obtain orientation history of HST prior to First Servicing Mission.




Link back to Hubble Space Telescope Hardware Archive Return to Hubble Space Telescope Hardware Archive System