Fusion Engineering and Design 18 (1991) 465-469 465 North-Holland

Some practical considerations in the design of in-vessel transporters

H o w a r d J o n e s

Spar Aerospace Limitea~ 1700 Ormont Drive, Westoh, Ontario M9L 2W7, Canada

Some design constraints common to various types of in-vessel tranlsporter are identified and discussed with reference to ITER. These constraints are determined by considering the effects of the transporter structural characteristics on the overall in-vessel remote handling system and the effects of the in-vessel environments on materials and components of the transporter. For structural design the importance of transporter natural frequency in addition to static deflection is identified. The tendency of transporters to exhibit a low-frequency torsional pendulum mode when handling heavy loads at large vertical distances from the equator of the vacuum vessel is demonstrated and its significance for remote handling is discussed. The effects of high ambient temperature on structural properties are evaluated with reference to the loss of strength and stiffness of aluminum alloys and the reduced load capacity of bearings based on Hertzian stress limitations. Potentially serious thermal de formation effects in non-iso-expansive structures are also briefly examined. Temperature constraints on the sizing of actuators are discussed, based on the allowable maximum temperatures of motor windings. It is concluded that these constraints are manageable provided they are recognized sufficiently early in the design process and some design guidelines for dealing with them are presented.

I. Introduction

The in-vessel remote maintenance of large tokamak machines requires a transporter to provide the major longitudinal displacements and coarse positioning of the remote handling system. Two basic types of in-ves- sel transporter (IVT) have been defined by various organizations: the articulated boom transporter (ABT) which is used at JET and TFTIL for instance, and the in-vessel vehicle transporter (IVVT) which is being studied for use in I T E R and FER. These two concepts are illustrated in figs. 1 and 2.

In spite of their differences in design and perfor- mance, which have been extensively discussed, both types of in-vessel transporter face some common prob- lems for which some general characterization and guidelines may be found useful. Certain of these com- mon problems are discussed below in two main cate- gories: - System Design: the question of stiffness - Equipment Design: the effects of temperature The former addresses the frequently-posed or implied questions: how stiff should the IVT be, how should its stiffness be defined and what are the critical cases? The latter reviews the effects of elevated temperatures on some common materials and components of in-vessel transporters and the implications for the equipment

design. While some general conclusions are made at the end of this paper, its main purpose is to stimulate further discussion and analysis of these issues.

Work Unit Articulated Boom

EqL Acc~

Main Support and

Insertion Assembly

Fig. 1. Articulated boom transporter.

E lsev ier Sc ience Pub l i she r s B.V.

466 H. Jones / Practical considerations of in-uessel transporters design

Equatorial Access Per

/ Divertor

Handling Unit

Fig. 2. In-vessel vehicle transporter.

Transfer Carriage for Divertor Plate

2. S y s t e m d e s i g n - t h e q u e s t i o n o f s t i f f n e s s

The IVT requirements typically specify that it should provide a rigid support base for the work unit. How- ever, no explicit stiffness requirements are usually de- fined. Generally it can be said that the stiffer the better from the point of view of the work unit since the opt imum case is when the work unit is mounted on an infinitely rigid base. However, to pursue this approach to its ultimate conclusion would result in an IVT design that would be unmanageable in terms of size, weight, and control. Furthermore, there seems to be no agreed definition of stiffness. Sometimes it is interpreted as static deflections of the IVT under its own weight and that of its payload, i.e. work unit and in-vessel compo- nent being transported. In other cases the static or dynamic response to small impulsive loads has been

considered.

The following approach was adopted by Spar for the purpose of a conceptual design of an in-vessel vehicle transporter with divertor handling unit: (a) Static or quasistatic deflections1 i.e. those tending

to remain approximately constant during a work unit movement, should not represent a problem for the work unit since it can compensate by a position adjustment. Hence, deflections due to weight of several hundred millimetres may be permissible.

(b) Dynamic deflections, i.e. those occurring within the time of a work unit function, are potentially disrup- tive and need to be carefully assessed in terms of amplitude and frequency. Since a large work unit used for divertor handling will not have the re- sponse to correct for dynamic deflections these should be kept low (2-5 ram).

(c) Natural frequency: even small displacements occur- ring near the bandwidth of a work unit can disturb its control system so ideally the natural frequency of the IVT should be well separated from the main control loop frequencies of the work units. Large work units such as the divertor handling unit ( D H U ) are expected to have a low bandwidth in the range 0.1-0.5 Hz so this implies that the IVT structural natural frequency should be approximately 10 times higher. Having a relatively high natural frequency also reduces the settling time of the system in cases where significant oscillations are induced.

The above are guidelines representing the simplified or "classical" approach to the design of a dynamic system. The corresponding results from structural anal- ysis of large in-vessel transporters are summarized in table 1 [1-3]. The mode shapes corresponding to the natural frequencies quoted in the table are illustrated in fig. 3. Table 1 is not intended to compare the merits of

Table 1 Typical structural characteristics of in-vessel transporters

Characteristic Concept

Articulated boom Vehicle

90 ° reach 180 ° reach Concept 1 Concept 2

Deflection (mm)due to 3700 kg weight 225

Deflection (rnm) due to 450 N lateral force 3

Natural frequency (Hz) with 3700 kg load - Vertical mode 1.7 - Lateral mode 0.6 - Pendulum mode 1.5

477 100 30

22 4 l

1.1 2.5 8.0 0.3 1.4 4.5 0.8 1.8 1.5

/ /

Vertical

H. Jones / Practical considerations of in-vessel transporters design

i i / ×

Lateral

d,// ~ Inverted Pendulum Pendulum

Fig. 3. Main mode shapes of in-vessel transporters.

467

3. I. Strength and stiffness of structural materials

Conventional structural materials, i.e. steel and aluminum alloys are expected to be used for the in-ves- sel transporters. More refined approaches to optimizing the structural properties of the IVT by the use of high-strength/high-modulus composite materials have been considered but are generally discounted because of cost and radiation resistance and the relatively small overall improvement obtained when the effect of the high extemal load of 3700 kg is taken into account.

Steel and aluminum alloys are negligibly affected by the radiation environment, even for long exposures. However, the effect of temperature on the strength and modulus of typical stainless steel and aluminum alloys are illustrated in table 2. Since with an ambient temper- ature of 150°C the expected total temperature of parts of the structure will be about 200°C the data shows that care must be exercised in the use of aluminum alloys because of their degraded properties. The effects on steel are less severe.

3.2 Load capacity and life of bearings

the various IVT concepts; rather it shows the expected performance of IVTs with variations, compared to the guidelines above. None of the IVTs can achieve natural frequencies as high as 5 Hz in all O f the modes of interest, although in general it is possible to obtain acceptable values of static and dynamic deflection. It is noticeable that the vehicle-type transporters are gener- ally stiffer than the articulated boom transporters as expected but that the performance of all design con- cepts is similar in the "torsional pendulum" modes. This mode is the main determinant of the system dy- namics approach for divertor handling. It implies that divertor handling operations must be accomplished using a low control system bandwidth, i.e. below 0.1 Hz.

Finally, it is recommended to avoid excitation of these low-frequency modes by using sinusoidal accelera- tion and retardation profiles during the manipulation of large loads.

3. Equipment design - the effects of temperature

The typical ambient temperature inside the vacuum vessel during maintenance is 150°C. This has certain implications for the design of in-vessel transporters which are discussed below.

The load capacity and life of rolling element bearings are generally limited by Hertzian contact stress and the resultant sub-surface fatigue. The allowable contact stress is determined by hardness which, for steel and aluminum, degrades with temperature to the same ex- tent as the tensile strength described in the former section.

However, the effects of temperature are more pro- nounced for bearings because of the contact stress and load-l ife relationships, e.g. for ball bearings: Contact stress cc (Load) °-33 i.e. Load Capacity cc (Hardness) 3. Also Life cc (Load Capacity) 3 for a given applied load. Hence Life cc (Hardness) 9.

Table 2 Tensile strength and modulus factors for metals at 150 °C

Material Tensile Young's strength modulus factor factor

AI alloy 2000 series 0.65 0.93 AI alloy 6061-T6 0.75 0.95 AI alloy 7075-T73 0.45 0.88 Stainless steel 300 series 0.85 0.95

468 H. Jones / Practical considerations of in-vessel transporters design

Table 3 Beating load capacity and life factors versus temperature

(a) Steel ball beating factors

Temperature ( o C) Hardness Load Life Capacity

25 1.0 1.0 1.0 150 0.97 0.90 0.73 200 0.95 0.86 0.64

(b) Aluminum roller track factors

Temperature ( o C) Hardness Load Life Capacity

25 1.0 1.0 1.0 150 0.67 0.45 0.07 200 0.48 0.23 0.01

Similarly for a cam roller on an anodised aluminum track it can be shown that: Load capacity cx (Hardness) 2 and Life cx (Hardness) 66.

The effects of temperature on load capacity and life are summarized in table 3. The results for a typical steel ball bearing, though more significant than might at first be supposed, are not severe since most bearings in the IVT are well derated in terms of load a n d / o r life

expectancy. For the case of the cam roller on an aluminum track

the results are quite dramatic because of the pro- nounced loss of hardness in the aluminum track.

3.3. Thermal deformation

The IVT will be assembled and tested at approxi- mately 20°C but will operate at a typical in-vessel temperature of 150°C. Over this temperature range the differential thermal expansion coefficients of the com- mon structural materials steel, stainless steel and aluminum alloy can have a significant effect.

While the overall structural deflections have a very limited or negligible effect on the alignment of the IVT, apparently innocuous differential thermal strains can cause problems for bearings because of their precision of assembly. This is illustrated in table 4 for a main bearing of the articulated boom transporter. Changes in bearing fit of 0 .2- to 0.4 mm are potentially serious since the alignment and internal clearance or preload of such a bearing are usually set with a tolerance of +0.1

m m .

Since it is not always possible to adopt an iso-expan- sive structure, i.e. an assembly of parts having the same thermal expansion coefficient, another solution is re-

quired. One convenient approach is to match the stiff- ness ratios of the bearing races to the housing and shaft such that the inner and outer races have the same effective expansion rates, i.e. the assembly becomes effectively iso-expansive for the temperature range of interest.

3.4. Performance and sizing of electrical actuators

The maximum rated temperature for electric motor windings is 220°C ( N E M A Class H), which is typically interpreted to mean that temperature predictions should be below 200°C. However, for most of these motors the ambient temperature is approximately 25°C so that power dissipations corresponding to a temperature rise of 175°C can be allowed.

For the in-vessel transporters only a temperature rise of 50°C can be permitted since the ambient is 150°C. This means that the power dissipations have to be correspondingly lowered, compared to the full rated values and implies that for a given load the actuator must be oversized. This is not a particularly attractive conclusion so other solutions have been considered:

- provide active cooling (gas or liquid), - allow winding temperatures above 200°C, - optimize the drive system for minimum power dis-

sipation. The first option suffers from the inconvenience of

routing the coolant supply to the many actuators in the system.

The second option is feasible (Spar has operated motors at 600°C in tests conducted for evaluation pur- poses) but should not be pushed too far. It should be remembered that the winding insulation is also being degraded by the radiation environment and that these motors must be very reliable because of the adverse consequences of an IVT failure inside the torus. If a careful review of the internal design of the motor is conducted, with particular attention to minimizing the damage caused by cyclic thermal expansion of the wind-

Table 4 Temperature effects on bearing fits

Material Coef. of Thermal Housing fit expansion strain change for ( / o C ) at150 °C 250mmOD

steel beating (mm)

Bearing steel 11 × 10 6 0.00143 0 Stainless steel 17x10 6 0.00221 0.20 AI alloy 23 × 10 6 0.00299 0.39

H. Jones / Practical considerations of in-vessel transporters design 469

ings, it is reasonable to allow a peak temperature of 250°C.

Thirdly, use of a high mechanical advantage, e.g. gear ratio, in the system, can reduce the required size and power of the actuator, since in general only very low output rates are required at the IVT. This is differ- ent from designing the drive system for maximum ef- ficiency since efficiency reflects the ratio of power out to power in rather than absolute values.

4. Conclusions

Some issues of general relevance for the design of in-vessel transporters have been discussed.

In the specification or evaluation of the stiffness of the IVT as a support base for work units, the impor- tance of dynamic displacements and natural frequency has been discussed together with a survey of results likely to be achieved by the various types of IVT. The need to consider all the main modes of deformation was emphasized by reference to the pendulum mode of the IVT when loaded with a divertor plate and handling unit at maximum vertical displacement from the equa- tor. The resulting constraints on the design of the work units was identified including some guidelines for ameliorating transient system dynamics effects.

The effects of the in-vessel temperature of 150°C on the equipment design was reviewed with reference to some main materials and components. In certain in- stances the solution to these effects takes the form of oversizing the components although it was shown that with careful design of the equipment the overall effect is not serious.

It is concluded that the issues discussed in this paper should not form serious impediments to the develop- ment of in-vessel transporters if recognized sufficiently early in the design process since the constraints are not excessive and can generally be accommodated by means of careful design.

References

[1] K. Waggitt, NET in-vessel handling unit concept step 3, Canadian Fusion Fuels Technology Project Report CF- FTP-9088-P. (December 1990).

[2] G. Clement, NET in-vessel vehicle system, Proof of Princi- ple Study of CEA Concept, Final Report, Commissariat a L'energie Atomique (January 1990).

[3] H.M. Jones, NET in-vessel vehicle system, Final Report for Concept Definition Phase, Canadian Fusion Fuels Tech- nology Project Report CFFTP-9021-I (December 1990).