Sloan Digital Sky Survey Telescope Technical Note 19970320
The Sloan Digital Sky Survey 2.5-m telescope has a uniquely large 3° field of view. The novel two-mirror optical design achieves zero distortion in the imaging mode using two transmitting correcting elements. However, to avoid excessive central obscuration, a conical baffle is necessary in addition to the usual primary and secondary baffles. This conical baffle is suspended approximately midway between the primary and secondary mirrors. It is formed of graphite fiber reinforced plastic because of the high stiffness to weight ratio of this material. The geometry of the part must be well-controlled. Consequently, it is formed on a machined aluminum mandrel.
The conical baffle was inspected on March 19, 1997 as it was being fabricated at Quality Composites, Inc. (QCI, Sandy UT). At that time, it was complete except several outside ribs remained to be attached and the upper and lower edges needed to be chamferred.
The conical baffle was the largest part of this sort ever attempted by QCI. Previously, the largest parts fabricated were straightforward tubes that were longer but smaller in diameter than the baffle. Three aspects of the baffle required development.
The final design consisted of eighteen graphite fiber fabric plies (Table 1 and Table 2). The hollow central stiffening rib was filled with polymer foam.
Table 1: The central stiffening rib was layed up in the following order. Ply Direction Comments 0°, 90° Fabric 0°, 90° Fabric 90° 90° 90° 0°, 90° Fabric 0°, 90° Fabric. Cure at 180°F for 11 hours under vacuum. None Foam. Cure at room temperature. Sand smooth. Table 2: The cone was layed up in the following order. Ply Direction Comments 90° 90° 0° Debulk at 120°F for 15 minutes under vacuum. 0° 90° 90° Cure at 180°F for 11 hours under vacuum. Sand smooth. 90° 0° 0° 90° 90° Cure at 180°F for 11 hours under vacuum.
Fabric
Fabric. Cure at 180°F for 11 hours under vacuum.
Foam. Cure at room temperature. Sand smooth.
Table 2: The cone was layed up in the following order.
Debulk at 120°F for 15 minutes under vacuum.
Cure at 180°F for 11 hours under vacuum. Sand smooth.
Cure at 180°F for 11 hours under vacuum.
Figure 1: The nearly finished baffle. Several outside ribs remain to be attached and the upper and lower edges need to be chamferred. The tape near the upper and lower edges mask surface that is not to be covered by the adhesive used to attach the fiber rib to the surface.
In the following discussion, the orientation of the baffle is assumed to be as shown in Figure 1, i.e., the baffle is setting on its large end with its small end up. This is the orientation that it will have in the telescope when the telescope is pointed at the zenith.
The outer rib locations were specified to be located within ±1.5 mm with respect to the inner ribs. This was checked by matching a full-scale plot to the part. Although it was difficult to measure the rib locations accurately with this technique, it was clear that the specification was satisfied and in most cases the rib locations were ±1.0 mm or better. Several outer ribs had not been installed at the time of the inspection and consequently were not measured.
The outer quarter of each outer rib was specified to be free of adhesive. It appeared that in one or two cases, this was not satisfied. QCI was asked to check the outer surface of each rib with a radius gauge after the inspection and to correct any ribs that did not pass this test.
The thickness of the cone was specified to be 1.5 ± 0.4 mm. The cone was measured at four equally spaced locations very near its upper and lower edge (Table 3). In the Tables, the angle 0° was at the split line between the two large mandrel sections and increased counterclockwise viewed from above the baffle. The angles measured coincide with the baffle attachment points. The mean thickness of the upper edge was 1.73 mm and that of the lower edge was 1.57 mm. All measurements were within the specification.
The height of the conical baffle was specified to be 726 ± 3 mm. Prior to measuring the baffle, it was placed upside down on a surface plate. A feeler gauge was used to determine that the upper edge was flat to better than 0.5 mm. The lower edge of the baffle was too large to measure on the available surface plate. Subsequently, the baffle height was measured at the same four angles (Table 3). The mean height was 726 mm. All measurements were well within the specification.
Table 3: The cone thickness near the upper and lower edges was measured at four locations around the part. The overall height of the part was measured at these locations also. Angle Upper edge thickness Lower edge thickness Height 0° 1.70 1.55 726.0 90° 1.85 1.60 726.0 180° 1.68 1.60 726.0 270° 1.70 1.55 725.5
Angle
Upper edge thickness
Lower edge thickness
Height
0°
1.70
1.55
726.0
90°
1.85
1.60
180°
1.68
270°
725.5
The diameter of the lower and upper edges of the baffle were designed to be 1239 and 996 mm respectively. QCI was concerned about the roundness of the baffle after it was cured. Consequently, a goal of 13 mm total indicated runout was set (±13 mm on the diameter). A tape measure was used to measure four diameters of the baffle at its upper and lower edges (Table 4). The mean diameters were 1239 and 998 mm. All measurements were within 3 mm of the design diameters. Although the measurement of diameters is insensitive to radius errors that vary as odd harmonics in angle, it would be most unlikely for any higher harmonic to be larger than the lowest even harmonic distortion. This is the familiar oval distortion.
Table 4: The diameter of the upper and lower ends of the cone was measured at four equally spaced intervals. Angle Upper diameter Lower diameter 0° 999.0 1240 45° 999.0 1240 90° 996.5 1239 135° 997.5 1237
Upper diameter
Lower diameter
999.0
1240
45°
996.5
1239
135°
997.5
1237
The estimated baffle mass was 9.43 kg but no tolerance was specified. It was measured on a balance to be 9.75 kg.
The quality of the conical baffle is excellent. All specifications were met or exceeded. In particular, the maximum minus minimum diameter was 3 mm or less for both the upper and lower edges of the baffle. The baffle has a fine appearance. The surfaces are smooth and uniform. The edges of the inside ribs are sharp and free of voids. Joints between ends of the acrylic fibers of the outer ribs are carefully made and visible only with careful inspection. Tests of the bond strength between the acrylic fibers and the surface give the excellent result that the fiber often fails before the adhesive.
Date created: 03/20/97 Last modified: 03/21/97 Walter A. Siegmund
