Summary

This study presents an evolution strategy integrated parallel solution algorithm for efficient size optimum design of large steel space frame structures, in particular high-rise steel buildings. An optimum design achieved for such systems is not of much practical value unless structural members are selected from a discrete list of ready sections according to strength, stability and serviceability requirements of a chosen code of practice. Besides, geometric compatibilities between beam and column members framing into each other at rigid joints have to be considered for practicality of the application. These points are observed in the solution algorithm developed such that the provisions of the Allowable Stress Design Code of the American Institute of Steel Construction (ASD-AISC [1]) are adopted for steel frames which are subject to vertical and horizontal loads, and standard wide flange hot-rolled sections are used to size the structural members. The algorithm has two novel attributes associated with the integrated optimization method and its parallel computing environment. It implements a very effective adaptive discrete evolution strategy method recently developed by Hasançebi [2]. The serial version of the solution algorithm, which allows for its implementation using a single processor, is first computerized using Borland Delphi source code. Next a master-slave configuration based parallel version of the algorithm is developed, where the sequential code is modified and reorganized with the advanced techniques of parallel processing to operate within a parallel computing system, which can be multiprocessors of a single computer or a cluster of computers connected via a local area network. The uniqueness of the parallel computing environment generated for the algorithm lies in the point that unlike FORTRAN and C++ programming languages, Delphi is not naturally supported by any MPI (Message Passing Interface) implementation. In order to implement MPI for the algorithm, special Pascal headers for MPICH2 [3] provided by the Lazarus Project [4] are revised according to the requirements of the algorithm.

The efficiency of the parallel algorithm in the optimum design of high-rise steel buildings is assessed using three examples. These examples are a 1040-member steel space frame with 60 member groups, a 3590-member steel space frame with 109 member groups and a 7648-member steel space frame with 198 member groups. The optimum designs are achieved for these frames using a cluster of computers with 32 processors in all. It has been shown that a maximum speedup ratio between 12.2 and 16.8 can been achieved for these examples. The performance of parallel computing system versus the number of processors employed is also examined in each example.

References

1 Manual of Steel Construction, “Allowable Stress Design”, 9th edition, AISC, American Institutes of Steel Construction, Inc, Chicago, Illinois, USA, 1989.
2 O. Hasançebi, “Discrete Approaches in Evolution Strategies Based Optimum Design of Steel Frames”, Structural Engineering and Mechanics, 26(2), 191-210, 2007.
3 MPICH2 Library, “High-Performance and Widely Portable Implementation of Massage Passing Interface (MPI) standard”, http://www.mcs.anl.gov/research/projects/mpich2, 2009.
4 Lazarus Project, “Visual Programming Environment for the Free Pascal Compiler”, http://www.lazarus.freepascal.org, 2009.

• A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems
• delphi and mpi, closer look 2
• DELPHI and MPI, closer look 1
• Delphi and mpi in 4 steps
• MPI Debugging with Visual Studio
• send std::string with mpi
• what i need from delphi 2010

The second part is about calling MPI functions from Delphi.

Name Convention

No change here. Since we are using a direct translation of the original MPI header, all the original names are there.

Sending, Receiving Data

No surprise here too, here are some pointers:

Variables

Since MPI needs pointers to variables, use @ operator to get the pointer to your variables.

MPI_Comm_rank(MPI_COMM_WORLD,@myid);

For getting the size of variables use the sizeoff function.

Arrays

Again we need the starting address of the arrays.

Lets say we have an array named as “sizeArray”

For dynamic arrays use @(sizeArray[0])

For static arrays use  @(sizeArray[1]) //  assuming the array starts from 1

Strings

Since no one knows the memory map of a Delphi string you cannot use them with MPI. You have to go with short strings.

Debugging

Bad news here, Delphi has no support for parallel debugging. You have to go with displaying data to screen, or using a log file.

I was going to write a 3rd article about creating parallel applications with GUI, but since it is a general subject I decided to write the article separately, so this series and here.

• Delphi and mpi in 4 steps
• Optimum Design of High-Rise Steel Buildings using an Evolution Strategy Integrated Parallel Algorithm
• DELPHI and MPI, closer look 1
• send std::string with mpi
• MPI Debugging with Visual Studio
• A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems
• pugg
• Installing boost for visual studio
• what i need from delphi 2010

In this first part I will take a closer look at the installation process.

1. Install an MPI library

MPI itself is not a library but an interface. So you have to find an implementation suiting your needs. MPICH2 is a long time favorite of mine, as it is free, well developed and documented.

All you have to do download and install the library on every computer of your cluster.

2.Get a Binding

Unfortunately, MPICH2 does not include bindings for Delphi. Fortunately Lazarus people created a pascal file  from the c include files that we can use.

You can find the slightly changed version of this file here Delphi mpi bindings (65).  All you need to do is download the file and insert it into your project. Since the dll files from MPICH2 will be in a path where Windows can find no other action is necessary.

3.Code!

Add uses MPI; to your code, that’s it. Here is a small example:

program MPIExample;
uses  MPI;

var myid: integer;
begin
  MPI_Init(nil,nil);
  MPI_Comm_rank(MPI_COMM_WORLD,@myid);
  writeln('id=',myid);
  MPI_Finalize;

end.

• Delphi and mpi in 4 steps
• Optimum Design of High-Rise Steel Buildings using an Evolution Strategy Integrated Parallel Algorithm
• delphi and mpi, closer look 2
• MPI Debugging with Visual Studio
• A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems
• send std::string with mpi
• what i need from delphi 2010

• Optimum Design of High-Rise Steel Buildings using an Evolution Strategy Integrated Parallel Algorithm
• delphi and mpi, closer look 2
• DELPHI and MPI, closer look 1
• Delphi and mpi in 4 steps

program HelloWorld1;

uses  MPI;

var myid: integer;
begin
MPI_Init(nil,nil);
MPI_Comm_rank(MPI_COMM_WORLD,@myid);
writeln('id=',myid);
MPI_Finalize;

end.

• DELPHI and MPI, closer look 1
• delphi and mpi, closer look 2
• Optimum Design of High-Rise Steel Buildings using an Evolution Strategy Integrated Parallel Algorithm
• send std::string with mpi
• MPI Debugging with Visual Studio
• A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems
• pugg
• Installing boost for visual studio
• what i need from delphi 2010