Ulrich A. W. Tetzlaff

Constraint Propagation Based Scheduling of Job Shops

This paper examines the minimum makespan problem for the job shop from an artificial intelligence point of view. A new solution procedure is presented which combines general learning strategies with special purpose heuristics, i.e., a constraint propagation based procedure is used and by decomposing the whole problem into single machine problems, problem specific knowledge is introduced. Fundamental for this approach is the equivalency between the constraint and the disjunctive graph. Presented comprehensive computational tests show high quality partial solutions, substantially accelerated exact methods, and a gain of deeper insight into the nature of the underlying problem instances. For example, the propagation and decomposition techniques reveal the particular structure underlying the well-known Fisher and Thompson 10 × 10 problem and thus provide an explanation for why this problem was intractable for such a long time.

Performance evaluation of flexible manufacturing systems with blocking

Real-life flexible manufacturing systems (FMS) often suffer to a considerable extent under utilization losses which are due to limited local and/or central buffer space and the limited velocity of the material handling system (MHS). In particular two forms of performance deterioration are distinguished: blocking and starving. A workstation is blocked if it cannot dispose of a finished piece because there is no buffer space — neither local nor in the central buffer. Blocking has been studied by several researchers (Akyildiz 1988a, 1988b, Altiok and Perros 1986, Suri and Diehl 1986, Gershwin and Schick 1983, Perros 1984, Yao and Buzacott 1985, 1986). In an earlier paper, the authors have made a proposal to estimate the influence of blocking on the performance of an FMS (Tempelmeier, Kuhn and Tetzlaff 1989) using standard closed queueing network algorithms. According to our experience, in industrial practice blocking often does not seem to be a severe problem during the configuration phase of an FMS, because in many cases it is possible to allow for enough central buffer space at moderate costs to prevent blocking. However even if the central buffer space is large enough, the limited velocity of the material handling system (MHS) in combination with limited local buffer space may lead to a form of reduction of machine utilization that is known as starving.

A queueing network model for flexible manufacturing systems with tool management

The analytical model in this paper allows the evaluation of the performance of a flexible manufacturing system (FMS) with a tool management system. Design parameters such as the transportation time for tools to machines, as well as the number of transportation vehicles for tools, are explicitly considered. The part and the tool transportation system are modeled as two interacting closed queueing networks. The classical convolution algorithm is used to evaluate the part transportation system and mean value analysis approximation is applied to evaluate the tool transportation system. The resulting set of nonlinear equations allows then to estimate important system parameters such as the throughput of parts, the utilization of the tool transportation vehicles and the service interruptions caused by a tool supply order.

Evaluating the effect of tool management on flexible manufacturing system performance

Abstract The design of tool management systems for Flexible Manufacturing Systems (FMS) is analysed. An algorithm is given, which allows performance evaluation of an FMS with its tool management system. The procedure applies queueing network theory in order to estimate important performance parameters. Of particular interest is the tool blocking time, which is defined as the time a machine is idle while it waits for the required tools.

Optimal workload allocation between a job shop and an FMS

Assuming given production requirements for a set of part types, we examine the question of how these production requirements can be assigned to different, alternative routes through two production systems, a flexible manufacturing system, and a conventional job shop. Several nonlinear optimization models are proposed in order to optimize performance parameters like throughput, work-in-process inventory, utilization, and manufacturing lead time. The models incorporate product form queueing network theory in order to evaluate system performance and extend results based on linear programming formulations. The solution procedures are based on decomposition, fixed point, and classical gradient search techniques.

A model for the minimum cost configuration problem in flexible manufacturing systems

This paper presents a mathematical programming model to help select equipment for a flexible manufacturing system, i.e., the selection of the types and numbers of CNC machines, washing stations, load/unload stations, transportation vehicles, and pallets. The objective is to minimize equipment costs and work-in-process inventory cost, while fulfilling production requirements for an average period. Queueing aspects and part flow interactions are considered with the help of a Jacksonian-type closed queueing network model in order to evaluate the systems performance. Since the related decision problem of our model can be shown to be NP-complete, the proposed solution procedure is based on implicit enumeration. Four bounds are provided, two lower and two upper bounds. A tight lower bound is obtained by linearizing the model through the application of asymptotic bound analysis. Furthermore, asymptotic bound analysis allows the calculation of a lower bound for the number of pallets in the system. The first upper bound is given by the best feasible solution and the second is based on the anti-starshaped form of the throughput function.

High cycle fatigue behaviour and generalized fatigue model development of lead-free solder alloy based on local stress approach

Abstract This paper gives an insight into high cycle fatigue (HCF) behaviour of a Pb-free solder alloy in the region between 104 up to 109 fatigue cycles using fatigue specimen. By means of a local stress approach, the method can be translated into solder joint fatigue evaluation in an application. The effect of temperatures (35xa0°C, 80xa0°C, 125xa0°C) on the fatigue property of Pb-free solder alloy is considered in this work to understand the possible fracture mechanisms and micro structural changes in a solder alloy at elevated temperature. Experiments are performed for different interaction factors under mean stresses (Rxa0=xa00, −xa01, −xa03), stress concentration (notched, un-notched) and surface roughness. SN (stress-life) diagrams presented in this work will compare the fatigue performance of Sn3.8Ag0.7Cu solder alloy for different conditions. Furthermore, mathematical fatigue model based on FKM guideline (in German “Fachkuratorium Maschinenbau) is extracted out of the experiments under all these external effects. The models can be exported later for lifetime evaluation purposes on applications. The paper thereby proposes the use of FKM guideline in the field of microelectronics.

Capacity optimization of flexible manufacturing systems under budget constraints

The capacity of a flexible manufacturing system (FMS) is optimized with the objective to maximize the systems throughput, while a budget constraint is considered. Decisions are performed on the capacity of machine groups (sets of identical machines), the transportation system and, in case of a significant cost impact, the number of pallets in the system. Throughput evaluation is achieved either by an open finite queueing network or by a closed queueing network if the number of pallets is included in the decision process. For both cases the solution procedure is based on the marginal allocation scheme.

Development of a TiC p Reinforced Ni-Based Superalloy MMC, with High Creep Resistance and Reduced Weight

Ni-based superalloys, in both single and polycrystalline varieties, are extensively used in high pressure turbine blades. But contrary to single crystal variants, the polycrystalline forms present easier manufacturing and offer higher potential for improvement in metal matrix composites (MMCs). To benefit from this opportunity, an Inconel X-750 superalloy reinforced with TiC particles is proposed, having a polycrystalline microstructure and the possibility for weight reduction in turbine elements application. The metallic powder with an addition of 15 vol.% of 3.7 μmd TiC particles was prepared through low energy mixing, uniaxial pressing and sintering, followed by a triple heat treatment. The microstructure was analyzed with SEM and XRD techniques. Compressive creep tests were performed at 800 °C with 200 MPa, on both original and reinforced alloys. The study shows how the inclusion of a highly compatible particle reinforcement does not only improves the creep resistance, but also reduces the material weight, thus having potential to promote further reduction in the creep rate on turbine blades submitted to centripetal forces.

Performance Evaluation of Repair Systems With Priorities

The paper describes a model for the performance evaluation of repair systems with priorities. Since for large problem sizes the calculations for the exact solution of our model are computational prohibitive, several heuristic solution procedures are presented. We investigate the quality of these heuristics with the help of several examples.