1. INTRODUCTION 1

2. FRAME 1

Appendix A

Figure 1: Solid model of 1/2 frame of original conceptual model.

Figure 2: Finite element deformed model of frame original conceptual model.

Figure 3: Finite element stress model of frame original conceptual model.

Figure 4: Solid model of 1/2 frame design with modified beams.

Figure 5: Finite element of deformed model of frame with modified beams.

Figure 6: Finite element stress model of frame with modified beams.

Appendix C

Figure 7: Angle check drawing for used ³He vacuum scattering chamber.

Figure 8: Assembly drawing of used ³He vacuum scattering chamber.

Figure 9: SHR smooth bore pump tee.

Appendix D

Figure 10: Physical layout of BLAST vacuum system, used for vacuum calculations.

Appendix F

Figure 11: BLAST assembly in closed position.

Figure 12: BLAST close-up of target area in closed position.

Figure 13: BLAST assembly with frame in split position.

Figure 14: BLAST close-up of target area in split position.

Figure 15: BLAST assembly split position, view from downstream end.

In this review, the concept of a split frame design is analyzed to prove feasibility. This design will allow target and inner detector access without breaking the ring vacuum. The design for the vacuum system for the internal target is also reviewed. The system was designed based on the worst case gas flow, which is 5X10¹⁸ atoms/sec (0.167 torr-l/s) for H₂ gas. Included in this review are the sizing and selection of the pumping system, a proposal for the design of the scattering chamber, beam line and wakefield suppressers. Also there is a budgetary cost comparison and analysis for the vacuum pumping system.

The conceptual model of the BLAST frame was analyzed to determine if the frame can be split into two halves along the beam line in order to facilitate access to the target, scattering chamber, and interior detectors. Specifications communicated orally included a requirement to not exceed a deflection of 2 mm in any member of the frame under full load. Another requirement was that the three octants ( 0° horizontal, ± 45° from horizontal on both sides of the beam line) had to remain clear areas for the installation of detectors. However, all of the area that is shadowed by the toroidal coils is usable for the support structure. The octants directly above and below the beam line (along the vertical axis) will be occupied by equipment and supports. A 3-D finite element model of the frame (Appendix A Figure 1) was made utilizing beam elements in the Ideas finite element program. The original conceptual design utilized 3" X 6" X 5/16" 316 SS tubing for most of the structural elements. When analyzed this frame had deflections in the 0.8" to 1" (20 to 25 mm) range and stresses were in the 11,000 psia range (Appendix A Figures 2 and 3) where the yield strength is 25,000 psia. The redesign mainly involved increasing the cross section of existing beams as well as slight manipulation of the location of key members in the frame (Appendix A Figure 4). This redesign also assumed three support rails on the floor to help distribute the gravity loads and minimize the size of the support beam for the bottom two toroidal coils. This was done so that the floor height beneath the BLAST does not have to be lowered. As a result of redesigning to the 2 mm deflection specification the stresses were extremely low. After several iterations of the design final numbers for a revised frame are 1.3 mm (0.051") deflection and von mises stress of 2,820 psia maximum (Appendix A Figures 5 and 6).

The large pumpout stacks, the turbo and roughing pumps with their electronics, and the cryogenic and laser systems will be supported by a mezzanine over the BLAST frame. The mezzanine will allow access to these systems without separating the two halves of the BLAST frame. This will give access to the vacuum system for maintenance, access to the cryogenics, access to the laser system for adjustments and maintenance, and access to the gas system for adjustment and maintenance. The cost of the mezzanine is estimated at $ 57,000 for engineering, design, materials and fabrication.

The modified design of the frame is a preliminary analysis to prove the concept of being able to split the BLAST frame with a minimum increase in material and cost. It is not a final design of the BLAST frame, it is only intended as a conceptual design of a modified frame to accomplish the deflection criteria. The cost of the modified frame is estimated at $207,000 for engineering, design, materials and fabrication.

The linear drive system for the BLAST halves will include linear roundway bearings and 2 inch shafts. We are also investigating a system using Hillman rollers in order to reduce costs and improve reliability. However for this study we used the Thomson costs to be conservative. The Thomson system will have a capacity of over 130 tons and be capable of moving the sections at a rate of one meter per minute. The drive system for each section will have a direct drive motor with 2020 oz-in torque and a recirculating ball nut on an acme thread and will have a capability of moving the section 1.8 meters. This gives a total space of 3.6 meters that will ensure adequate room to access the central components in BLAST. Approximate cost for the drive system components is $15,700 per half or approximately $31,400 total. Figures in the appendix B show the numbers and assumptions used in sizing the system and the list of components and their costs. The excess load capacity of the system is due to the extra support needed because of the lack of stiffness in the BLAST frame due to height restrictions on the members composing the bottom of the frame.

The scattering chamber has several requirements to meet the specifications of the ring as well as the BLAST experiment. The chamber must be made to fit inside the BLAST toroidal coils including all services, (cryogenics, laser pumping cell, etc.). The chamber must have the appropriate opening to allow for all the angles that the experiments require. These include 17° in-plane forward angle to 90° in-plane back angle and +/- 20° out of plane angle, all these numbers being from the interaction point (Appendix C Figure 7). To ensure compatibility with the ultra-high vacuum of the SHR all components will be machined and cleaned according to "good" vacuum practice. This includes using only hydro-carbon free cutting fluids and solvents. All seals for the system will be made using metal seals. This includes the aluminum flanges which will utilize Altine TiN coated aluminum conflats using aluminum gaskets. All components of the system will be capable of repeated bake-outs to 150° C. The system must also be "non-magnetic" having materials that cannot be excited to magnetism. This means an all aluminum chamber and seals with the bolts and threaded inserts being the only steel (stainless steel) components inside the toroidal field. The bolts will be specified to be 316 SS and the threaded inserts will be 304 SS. This will minimize their contributions to perturbations in the magnetic field. Finally the chamber must be compatible with the target. This implies that different chambers will be needed for the different target gas sources. The laser driven sources can be accommodated by one chamber type and the Atomic Beam Source will require another chamber type. For the laser driven sources an existing chamber used in a previous ³He experiment will meet all of the above requirements (Appendix C Figure 8). New flanges and supports will need to be designed and manufactured for this chamber to be used in the BLAST spectrometer. Included in these designs will be cryogenic cooling capability which was not used with this chamber previously. This chamber is currently on campus and is available. A second chamber will have to be designed and manufactured for the Atomic Beam Source. Estimated costs for this chamber are $28,800 for engineering and drafting and $22,500 for manufacture plus the cost of any transfer lines from the source to the chamber.

The vacuum system must be capable of pumping the severe gas load created by these internal targets. The gas flow is estimated at 1X10¹⁸ atoms/sec for ³He and 5X10¹⁸ atoms/sec for H₂ and D₂. These correspond to rates of 0.033 torr-l /sec and 0.1667 torr-l /sec respectively. The nominal ring vacuum is in the 1X10^-9 torr range and it will be necessary to be in the 1X10^-7 torr range in the target section in order for the ring vacuum to remain at it’s nominal level. The system also needs to be a "dry" pumping system so as to not introduce contamination especially hydro carbons. The only components capable of pumping this amount of gas for an extended amount of time while maintaining a "dry" environment are turbo pumps backed by "dry" roughing pumps.

Another consideration is rf wakefield heating from the beam. This means that the beam should perceive a continuous wall as it travels through the interaction area. This is accommodated by wakefield suppressers, perforated tubes that allows gas to get to the vacuum pumps but has enough material remaining to have the appearance of a continuous wall to the electron beam. The perforated tube will be made of a 316 SS mesh that is 1/32" thick with 3/32" diameter holes and a 50% open area. This design utilizes the same basic design of current pumpouts used in the SHR, a drawing of the SHR pumpouts is included in the Appendix D Figure 9.

The proposed system design is based on the previous considerations as well as vacuum calculations included in Appendix D and experience from the HERMES experiment at DESY. The gas load contribution due to outgassing, permeation and leakage of the vacuum system is ignored because it is orders of magnitude lower than the source gas ( ³He and H₂ ) contribution. The system (Appendix D Figure 10) will include six 1000 liter/sec turbo pumps backed by either a hook and claw or scroll roughing pump. Diaphragm pumps were eliminated from consideration because of their tendency to retain helium, making leak checking impossible. The same size pumps were chosen to give part redundancy so that a spare of only one type of pump is required. This will give a vacuum level of 8X10^-8 torr at the end of the pumping section for ³He. For the H₂, because of the larger flow rate, another stage of pumping will be added at the end of the turbo pumping section . This section will contain a getter wafer module that has a pumping speed of 770 l/s. This will bring the vacuum level, for Hydrogen gas flow, down to 2X10^-7 torr or less. These numbers meet the verbal specifications as given by Ernie Ihloff of the Bates Laboratory Staff.

Several manufacturers were consulted and asked to give budgetary quotations for the above pumping system. The manufacturers and their systems and quotations to date are included in the Appendix E. For the "dry" backing pumps there are fewer manufacturers that meet our requirements. The main pumps under consideration are the Edwards Drystar and Ebara. The Edwards Scroll pump has reliability problems and the Kashiyama has no support network in the United States. Both the Drystar and Ebara require cooling water and a nitrogen purge. These pumps have far in excess of the backing requirements for one pump, so it is planned to manifold three turbo pumps per roughing station. This will justify the high cost of the roughing stations, approximately $30,000 per station.

In Appendix F there are several figures (from different views) representing a solid model of the modified BLAST frame with the mezzanine and vacuum system. Not included in these drawings are supports for the beam line and the linear drive system. They will be included in a later version once the present plan is adopted. This solid model is only a conceptual design although many of the components closely represent the components of the final design.