Torsional Stiffness Measuring Machine (TSMM) & Automated Frame Design Tools William Thomas Steed Sept. 8, 2009 Bachelor of Science in Mechanical Engineering Masters of Science in Mechanical Engineering College of Engineering Committee Chair: Randall Allemang
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Designing an automotive chassis is not an intuitive process. It, at times, can be
very difficult depending on the geometry of the structure. Research was conducted at the
University of Cincinnati to alleviate the burden of this task. Software tools were
developed to help speed the design process. A new technique of measuring the torsional
stiffness of a Formula SAE chassis design was created. Finally, a recommended process
is presented to perform the design and validation of a Formula SAE chassis.
As engineers we turn to different tools that we have access to in order to
understand and iterate a design. In the area of space frames, design tools can be limited.
To get an understanding of a chassis design, engineers turn to Finite Element Analysis
(FEA) to gain a better understanding of these types of structures. Ultimately, manual
iterations are not enough to completely optimize a structure to a desired goal. Software
tools need to be developed in order to have a deep understanding of how the structure
performs at each iteration. Two tools, a sensitivity and optimization tool, were written
and the out come of each is discussed.
Until 2007, the UC Formula SAE team has validated only the current years frame
design and not the entire chassis design. In the world of Formula One racing it is
essential to have knowledge not only of frame stiffness but also hub to hub chassis
stiffness. Various ways to test chassis stiffness were investigated and designed. A static
test was developed and performed. A finite element model and its correlation to this
static test is discussed.
iii
COPYRIGHT NOTICE
iv
Acknowledgement
This thesis has been the greatest culmination to an engineer-in-training process. I have great admiration for the following people because of their eagerness to help, their ambition to learn and their patience to listen and mature my ideas.
Thank you:
• To the “man upstairs” for giving me all the wonderful blessings of this life and sharing with me in all these years.
• To my family for always believing in me, providing for me and giving me inspiration to be the best engineer I know how to be.
• To the University of Cincinnati for providing a faculty of the best engineering professors and learning facilities to chase a student’s dream.
• To Doctor Randall Allemang, for giving me the freedom to explore my ideas and for being a superb engineering role model.
• To Doctor Allyn Phillips for aiding in the development of my Matlab skills and for your patience while I shared the lab’s equipment.
• To my thesis committee. Thank you for your thoughts and time.
• To Douglas Hurd and David Breheim. Thank you for expert advice and patience.
• To the 2005 - 2007 University of Cincinnati FSAE teams for sharing your ideas, your talents and your passion for building race cars and believing that this research can provide a deeper understanding of each design.
• To my colleague Benjamin Stoney; without your help this would not have been possible. Thanks brother, for working as hard on these cars as you do and for all the great welds.
• To my colleague Fredrick Jabs for conversing with me to mature my ideas and pushing the limits of engineering design.
• To my colleague Ryan Lake for setting great examples for future teams and engineering students. Thanks for your thoughts and time. It has been fun!
• To my colleague David Moster for all the long loud years of learning how to become great engineers. Thanks for keeping us fast!
• To my colleagues Ben Rawe, Abbey Yee, Ravi Mantrala and Bill Wise for the extra thoughts and hands they provided during testing.
• To my colleague Dan Alford for your support and dedication to getting the University of Cincinnati FSAE back to top 5.
• To Carroll Smith for creating a collegiate activity that challenges engineers to be better than ever could have thought they could be. Preparing for and competing in this series has been the one of the greatest accomplishment of my life.
This thesis is dedicated to my family and friends: Margaret and Ray Winialski, William, Brian, Kathleen, Edward & Jean Steed, Robert Boehm, Sara, Grant, & William Leto, Mary Ann & Norman Noe, Josh Kullis, Sindney Tippet & Paul Tinetti.
v
Table of Contents List of Figures .................................................................................................................... vi
List of Tables ............................................................................................................... viii Chapter 1 Development of the Race Car Frame ................................................................. 1
The Ladder Frame ........................................................................................................... 2 The Space Frame............................................................................................................. 3 The Composite Monocoque ............................................................................................ 4
Chapter 2 ............................................................................................................................. 7 Chapter 3 Frame Model .................................................................................................... 11
Geometry Construction ................................................................................................. 12 Element Types .............................................................................................................. 14 Material & Section Properties ....................................................................................... 16 Meshing......................................................................................................................... 18 Frame Model Design Iterations & Constraints ............................................................. 19
Chapter 4 Chassis Model .................................................................................................. 24 Why Model the Chassis? .............................................................................................. 25 Revolute Joint ............................................................................................................... 30 Model Constraints ......................................................................................................... 33
For Formula SAE teams it is not only a race to the finish line, but also a race to
finish the building and validation of a racecar. Having every tool possible to “get there”
gives teams the upper advantage. The design and validation of the chassis as a whole
will continue to be the focal point at every collegiate racecar competition. Understanding
and utilizing the tools presented in the preceding will not only speed the process, but will
also provide that extra edge for competition.
The process from frame to chassis model is an engineering process that has been
proven to work at the University of Cincinnati. Establishing and documenting this
process will allow future teams to achieve the goals they have for their designs.
Following this process also allows for a direct comparison from year-to-year. When rule
changes are so widespread that a year-to-year comparison can not be made, engineers can
find reassurance that if this process is followed, the result will be a very good working
product. This process gives excellent trend (relative) information to evaluate progress as
well as accurate absolute (properly scaled) numbers for ultimate design and analysis.
The automation of the sensitivity analysis played a key role in the success of this
process for the UC Formula SAE team. Having a global view of how each tube’s
stiffness contributed to the overall stiffness of the frame was a huge step forward in the
design process. The ability to quickly make changes as manufacturing issues arose was
an invaluable asset. The sensitivity analysis tool saved time by pointing the designer in
the direction that needed to be taken in order to improve frame stiffness. In the end, this
allowed for more time to be dedicated to the building and manufacturing and assembly of
62
the vehicle. The sensitivity analysis proved not only to be a great tool for this type of
design work but it also has the potential to be adapted to aid in the design of any
structure.
In the end, the optimization tool presented did not turn out to be the ultimate
design tool. The number of combinations needing to be solved in such a short period of
time overwhelmed the idea. However, the results of the small subset that was evaluated
showed great potential. To have absolutely no human interaction required to complete
the design of the frame would be a tremendous feat of ingenuity. It would, without a
doubt, give any team the upper hand. With advancements in CPU speed and a more
robust code, this tool could quickly become a very valuable engineering tool.
Testing and validation is an important part of any engineering process. Testing
allows engineers to prove not only to themselves but to others, that what was designed is
that which was built. Having a test rig such as the TSMM allows for a direct comparison
from model to hardware. Comparing the data from the rig to analytical data showed that
the correct loading scheme was achieved. The results of torsional stiffness showed that it
can produce very accurate answers. This gives confirmation that the actual hardware will
perform as it was designed to. Results boost team confidence and gives reassurance that
the chassis can make it to the finish line. Having this testing process established plays a
large role in communicating with design judges in a collegiate competition. The ability
to relay that a $500,000 four post dynamic shaker system was modified to statically
measure chassis torsional stiffness is an impressing and interesting conversation starter.
Having a test rig like the TSMM, shows design judges that your design innovation does
not just stop at the car, but is continued into accurate testing and validation methods.
63
This also justifies that future designs can proceed analytically with only minimal testing
and validation.
Design Tools
Beyond the design tools presented there is much more to consider. A script could
be developed to automatically take a Solid Edge model and turn it into a working
ANSYS frame model. The leap between the frame model to the chassis model would be
much more difficult. With the help of a Matlab GUI, it could be achieved. The notion of
having a set of points loaded into a Matlab GUI and then formatted into an ANSYS
database is very possible. The result would yield an entirely automated process, start to
finish, eliminating the need for the more complicated processes outlined in Chapters 2
and 3.
If the choice to completely automate the analysis process from Solid Edge to
ANSYS is not chosen, the recommended path is to make the solution routines a little
more user friendly. This could be completed with the use of a feature in ANSYS. The
scripts can be added in as a command in the toolbar. The result would be such that each
time the ANSYS GUI is opened, the commands to run the scripts are only one click
away.
The most difficult tool to write would be a geometry optimizer. The scripts
presented in this text do not take into account that, between the defined points, the
arrangement of tubes could be different. A script could be written to identify whether it
is more important for a frame member to be oriented at one angle versus another. A
script of this nature could be derived from the sensitivity routine. Components could be
64
defined such that at times they are meshed and at others, are not. This would measure the
difference in torsional stiffness with and without this component in the frame.
Additionally, the link between modal analysis and experimental modal analysis
could be better defined with the help of a cross-orthogonality script. This script would
extract the modal vectors (mode shapes) of the analysis and do the same for the
experimental data. These two vector pairs can then be compared for each mode and a
better understanding of how well the analysis actually compared to experimental data.
Once the set of vectors are defined for each mode they can be compared for how parallel
they are. This would be the ultimate modal analysis correlation tool.
The Taguchi Method could also be implemented to help optimize the frame. The
Taguchi Method was developed to optimize manufacturing processes, however it could
be adapted to perform case studies on the frame. The Taguchi method utilizes orthogonal
arrays. Orthogonal arrays are tables which define a minimum number of experiments to
be able to understand the importance of one parameters versus another. When the data
has been collected, signal to noise ratios are calculated for each parameter. The
parameters with the largest signal to noise ratios are then considered more important [4].
Another optimization process that could considerably improve the frame design
would be to implement the use of objective functions. A function could be developed,
along with a set of inequalities, that when solved using differential calculus could
generate the optimal frame design [15]. An example problem using objective functions
has been illustrated in Chapter 2 of Reference 15. This optimization technique was also
found to be in current Finite Element (F.E.) Programs and should be investigated to
determine its value in improving future frame designs.
65
Torsional Stiffness Measuring Machine
The TSMM was designed and built because it is extremely versatile and possesses
opportunities for improvement. There are some basic upgrades that could be made and
some extensive ones. All of these upgrades should be completed to better understand the
ability of the chassis to do its job and perform at its peak.
To begin, there are various minor modifications that should be made to the rig.
First, the holes that were created in the base of the each fixture were reamed to the size of
the fixture bolts. These bolts when purchased were bowed. This did not allow for a very
easy installation. There are two options that could be executed that would easily remedy
this situation. Better grade bolts could be purchased or the sixteen holes could be reamed
out with an over-sized reamer. Knowing that the bolts of this type, length and grade are
expensive, the recommended way is to use the over-sized reamer.
The solid shocks designed for the TSMM were difficult to use because the correct
rod ends were not available. Therefore, thin spacers needed to be created to hold the
center-line of the solid shock at the correct height. Trying to install two spacers above
and below the rod end inside the frame pickup tube presented a frustrating challenge. For
the future, it is recommended that, each year, a custom set of solid shocks be made at the
correct width of the shock mating interface. The other option would be to make a
tapered sliding dovetail joint that could be adjusted for length and locked at the correct
height. Depending on the construction of this joint, it may pose problems while testing.
The dovetail could potentially slip if it is not pinned. The recommended way to fix this
issue is to make a custom set of solid shocks. Each year, a solid shock assembly should
66
be an automatic item to be fabricated by the suspension team. This solid shock assembly
should be designed to be integrated into the TSMM.
To increase the confidence in the results of the TSMM data, an easy upgrade
would be to acquire some steel pins to replace the “fuse” in the fixture. The solid brass
fuses performed without any indication of yielding. However, the fuses were under
constant load when vehicles were attached to the rig. Replacing these fuses with a
material that has a higher shear modulus would be recommended.
For testing, a great addition to the TSMM would be to connect the two forward
posts to move in unison with one button press of the controller. This would eliminate the
pause from the movement sequence of each front post. Inherently, this would eliminate
the need to identify the correct time period in which both post have been moved. In order
to accomplish this, internal coding in the hydraulic controller software may be required.
This would probably involve some heavy “C” coding and would take quite a few months
to integrate. A better solution would be to go back to the software design group and ask
if this option could be added into the next revision of the software.
The TSMM was conceptualized and built to have many more capabilities than just
a torsional stiffness measuring rig. The rig is capable of being setup to test suspension
parameters. If a table was designed and built to hold the frame fixed, a single post could
be translated while suspension parameters are measured. This type of testing eliminates
the use of solid shocks and poses less harm to the car. This test would replace the need to
travel to Goodyear to utilize their suspension parameter measuring machine and would
give the Formula team that much more time to understand the current suspension design.
Having built and designed a suspension parameter measuring machine would also
67
provide more justification that the Formula Team understands the suspension and how it
is suppose to perform, when presenting to the competition judges. The rig would need
some modification to accomplish this. A three point sensor system would need to be
developed. This could be done with the use of string potentiometers or perhaps a laser
tracking system. In addition, a kinematic table would need to be developed to pitch and
yaw the frame. The cost of materials and sensors to create this addition would probably
overwhelm the budget and consume valuable design time. The system would take many
years to develop. The software alone to process and create the curves would require the
suspension team leader’s time and some heavy Matlab code training.
The more desirable addition would be to use the TSMM rig as a quick way to
simulate on track conditions if weather or track availability presented issues. A Motec
data acquisition system was purchased in 2007 for acquiring on track data. Information
from the Motec data acquisition system could be used to produce a time history of
displacement at each wheel. This data could be formatted as input into the four post
simulation software. With the input established the simulation could be run and team
members could closely monitor chassis performance. During the simulation, strain gauge
information could be logged on any part of the vehicle. This would also provide the
chance for team members to check operating clearances, which could warrant design
changes and prevent hardware failure.
68
Figure 55: Commercial Semi-Dynamic Test Rig [9]
Figure 55 shows a commercially available rig to do this type of testing. In this
instance a Nextel Cup series chassis has been installed and track conditions are simulated.
The rig is able to translate in all three translational degrees of freedom. This poses a
problem for the TSMM, but could be remedied. The four post MTS system of the
TSMM was not designed to allow for in-plane translations. The posts have an air ride
system that makes moving them much easier, but it requires a large demand for air. The
air supply system could be upgraded to allow for all four post to be run at the same time.
The other option is to place two steel plates underneath each post with an oil film in
between. This is a relatively simple solution if oil is applied manually. A more
permanent solution would be to use granite slabs attached to the floor with a steel top
plate and an oil pump. This system if built properly, would be fairly easy to maintain and
would be dependable. In any case, the first two solutions would be more economical.
69
References
1. Aird, Forbes. Race Car Chassis: Design and Construction. Osceola:MBI, 1997 2. Sakkis, Tony. Anatomy & Development of the Indy Car. Osceola: MBI, 1994 3. Weissler, Paul “Body by the numbers.” Automotive Engineering International Sep. 2007: 26+ 4. Roy, Ranjit. A Primer on the Taguchi Method. Dearborn, MI: Society of Manufacturing
Engineers, 1990 5. Beer, Ferdinand, E. Johnson, and John DeWolf. Mechanics of Materials. New York, NY:
McGraw Hill, 2001.
6. Perry, C., and H. R. Lissner. The Strain Gage Primer. New York, NY: McGraw Hill, 1962
7. Student Manual for Strain Gage Technology. Raleigh: North Carolina, 1992
8. Strain Gage Conditioner and Amplifier System Instruction Manual. Raleigh: North Carolina:
TD, 1992.
9. Monaghan, Matt “New Simulation Rig Help ‘Tighten’ Racecar Performance.” Automotive Engineering International Dec. 2007: 62-63
10. Adams, Herb. Chassis Engineering. New York, NY: HP Books, 1993 11. ANSYS Academic Research, v. 8.0 12. ANSYS Academic Research,8.0, Help System, Command Reference, ANSYS Inc. 13. ANSYS Academic Research 8.0, Help System, Element Reference, ANSYS, Inc. 14. Solid Edge Academic ,V17 15. Haftka, Raphael and Zafer Gϋrdal. Elements of Structural Optimization. Dordrecht,
Netherlands: Kluwer, 1992
63
Model Dimensions Sufficient Mesh Density No Cracks/Discontinuity No Distorted or Warped Elements Correct and Consistent Units for Properties/Loads Element Type Keyopts Component Separation with Gap Elements Temperature (TREF, TUNIF, and applied temps) Heat Transfer Coefficients and Gas Temperatures Displacements and Restraints Couples Constraint Equations Pressures Forces and/or Moments Rotational Velocity and/or Accelerations Rotated Nodes Large Deflection
Stress Stiffening
Output Check Reaction Forces Check Mass vs. Supplied Data or Hand Calc’s
View and/or Plot Model by
Element Types Real Constants Material Properties Boundary Conditions
View and/or Print Listing by
Element Types Real Constants Material Properties
Apply Unit Displacement Apply Uniform Temperature Apply 1G Acceleration Modal Analysis (Modes and Frequencies) Deformation Check Flexibility Check
Jobname: ________________ Solved Using: (ANSYS Ver. __________ Other_________________________) # of Nodes: ______________ Analysis Type: ( Static Modal Thermal Other__________________) # of Elements:____________ Equation Solver: (Frontal PCG Block-Lanczos Other____________) # of D.O.F.:______________ Hardware Used: ( Pentium HP Sun Other______________________) Filesize (.db):_____________ Memory Required for Run (Total/Database):_______________/______________ Filesize (.rst):_____________ Disk Space Required for Executing a Run:_______________________________ Units: (English Metric) Run Time (CPU/Clock): _______________________/______________________ Source of Geometry (Drawing #, CAD filename): ______________________________________________________ Software used for Mesh Generation: (ANSYS PATRAN I-DEAS Other____________________________) Model Description: _______________________________________________________________________________ ________________________________________________________________________________________________ Loading Description: ______________________________________________________________________________ ________________________________________________________________________________________________ Restraint Description: _____________________________________________________________________________ ________________________________________________________________________________________________
Model Checks Input Check Runs
64
Appendix B Scripts Torsional Stiffness Script ! tstiff.txt ! W.T. Steed ~ 5/14/07 ! University of Cincinnati ! EMAIL: [email protected] ! TEL:(513) 260-8955 ! !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- ! PURPOSE: Calculates the torsional stiffness of an SAE ! Frame ! OUTPUT: Torsional Stiffness to background screen and to ! tstiff variable !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- ! !******************************************************************** ! Notes: ! - This script is meant to be used with a Frame Model in ANSYS ! - Meant be run INTERACTIVELY ! - Assumes that CYLINDRICAL coordinate systems ! 200-205 have been created at the loading and restraint ! locations on the frame !******************************************************************** !******************************************************************** ! Step 1: Define Track Width !******************************************************************** couple_length=17/12 ! Input in feet !******************************************************************** ! Step 2: Define Restraint Nodes !******************************************************************** !d1=7158 !d2=7190 !d3=7202 !d4=7236 csys,200 nsel,s,loc,x,0,0.1 *get,d1,node,0,num,min csys,201 nsel,s,loc,x,0,0.1 *get,d2,node,0,num,min csys,202 nsel,s,loc,x,0,0.1 *get,d3,node,0,num,min csys,203 nsel,s,loc,x,0,0.1 *get,d4,node,0,num,min !******************************************************************** ! Step 3: Define Force Nodes and Value !******************************************************************** fval=100 !f1=6614 !f2=6643 csys,204 nsel,s,loc,x,0,0.1 *get,fc1,node,0,num,min csys,205 nsel,s,loc,x,0,0.1 *get,fc2,node,0,num,min
!******************************************************************** ! Step 4: Define what node to take displacement from to ! calculate the angle the frame twisted !******************************************************************** dnode=fc1 !******************************************************************** ! Step 5: Apply Boundary Conditions
Sensitivity/Eigen Value Analysis Script ! tstif_eigen_analysis.ain ! W.T.Steed 11/20/06 ! University of Cincinnati ! EMAIL: [email protected] ! TEL:(513) 260-8955 !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- ! PURPOSE: ! - Calculates Torsional Stiffness and Computes first 6 Modes ! - To track how modes of vibration change with tor. stiffness ! OUTPUT: EIGEN.TXT !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- !******************************************************************** ! Notes: ! - This script is meant to be used with a Frame Model in ANSYS ! - It should be run in BATCH Mode ! - It assumes that CYLINDRICAL coordinate systems 200-205 ! have been created at the loading and restraint locations on the ! frame !******************************************************************** !******************************************************************** ! Instructions: ! 1. Create a component for each pair of symmetric members in the ! FRAME Model ! - Name each component "line1, line2, line3, etc." ! - Note how many components there are ! ! ******Make sure that there are no lines from the engine in any of ! these components***** ! ! 2. Make sure there are 17 sections defined, ensure that section ! 17 is the same as section 1 ! ! 3. Save a copy of the Database ! ! 4. Modify the /filn command to be /filn,"database name" ! ! 5. Modify the number 39 with the number of the "line" ! components in the database ! ! 6. Execute in Batch Mode and watch check EIGEN.txt every so ! often !******************************************************************** !******************************************************************** ! Step 1: Read in Database !******************************************************************** /batch /filn,07-tstiff-eigen resu !******************************************************************** ! Step 2: Define initial parameters !******************************************************************** /prep7 *get,max_secp,secp,num,max *get,count_line,line,0,count *DIM,eigen_stiff,array,max_secp,7,39 !******************************************************************** ! Step 3: Begin Solution Routine !******************************************************************** *DO,ii,1,39,1 /prep7
Page 1
allsel allsel,below,volu lsel,inve ! Set Section number and material to begin the first mesh with secnum,1 mat,1 cmsel,all ! Loop through each cross section *DO,jj,1,17,1 cmsel,s,line%ii% /prep7 lclear,all latt,,,,,,,jj secnum,jj lmesh,all couple_length=17/12 csys,200 nsel,s,loc,x,0,0.1 *get,d1,node,0,num,min csys,201 nsel,s,loc,x,0,0.1 *get,d2,node,0,num,min csys,202 nsel,s,loc,x,0,0.1 *get,d3,node,0,num,min csys,203 nsel,s,loc,x,0,0.1 *get,d4,node,0,num,min fval=100 csys,204 nsel,s,loc,x,0,0.1 *get,fc1,node,0,num,min csys,205 nsel,s,loc,x,0,0.1 *get,fc2,node,0,num,min dnode=fc1 /prep7 allsel d,d1,all d,d2,all d,d3,all d,d4,all ! Force f,fc1,fz,fval f,fc2,fz,-fval /solu antype,0 allsel solve /post1 set,1 csys,0 *afun,deg *get,dnode_u,node,dnode,u,z *get,dnode_y_loc,node,fc2,loc,y theta=atan(dnode_u/dnode_y_loc) tstiff=((couple_length*fval)/theta)
Page 2
66
! Store torsional stiffness in first column of eigen_stiff array eigen_stiff(jj,1,ii)=tstiff ! Clear Boundary Conditions /prep7 lsclear,all ! Start Modal Analysis /solu antype,modal MSAVE,0 MODOPT,LANB,12 EQSLV,SPAR MXPAND,0, , ,0 LUMPM,0 PSTRES,0 MODOPT,LANB,12,0,10000, ,OFF solve /solu ! Loop through Modes 7 to 12 and store in colums 2 ! through 7 of eigen_stiff *DO,kk,2,7,1 xx=kk+5 *get,freq,mode,xx,freq eigen_stiff(jj,kk,ii)=freq *ENDDO *ENDDO /prep7 ! Save all Parameters to eigen.txt PARSAV,ALL,'eigen','txt',' ' allsel *ENDDO
Page 3
67
Create Combo Script ! create_combo.txt ! W.T.Steed ! University of Cincinnati ! EMAIL: [email protected] ! TEL:(513) 260-8955 !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= ! PURPOSE:- Creates Matrix of all possible combinations for optimization ! OUTPUT: COMBO Matrix in ANSYS Parameters !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= !******************************************************************** ! Note: ! 1. Takes approximately ten minutes to execute !******************************************************************** !******************************************************************** ! Instructions: ! 1. Run interactively in any ANSYS database ! 2. Save Database ! 3. Run Optimization Routine !******************************************************************** ! Dimension the "COMBO" Matrix *DIM,COMBO,array,65536,4,1 *DO,i,1,4096,1 *DO,j,1,16,1 x=(i-1)*16 COMBO(x+j,4)=j *ENDDO *ENDDO counter1=0 counter2=0 counter3=0 dummy1=1 dummy2=1 dummy3=1 *DO,i,1,65536,1 counter1=counter1+1 *IF,counter1,eq,17,then counter1=1 dummy1=dummy1+1 *ENDIF *IF,dummy1,eq,17,then dummy1=1 *ENDIF COMBO(i,3)=dummy1 counter2=counter2+1 *IF,counter2,eq,257,then counter2=1 dummy2=dummy2+1 *ENDIF *IF,dummy2,eq,17,then dummy2=1 *ENDIF COMBO(i,2)=dummy2
Optimization Script ! Optimization.txt ! W.T.Steed ! University of Cincinnati ! EMAIL: [email protected] ! TEL:(513) 260-8955 !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- ! PURPOSE: ! - Optimize 4 components of the Frame Model ! OUTPUT: Paramaeters.txt !-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- !******************************************************************** ! Notes: ! - This script is meant to be used with a Frame Model in ANSYS ! - It should be run in BATCH Mode ! - It assumes that CYLINDRICAL coordinate systems 200-205 ! have been created at the loading and restraint locations on the ! frame ! - This analysis takes multiple weeks ! - Analysis blocks may need to be split into smaller block sizes ! i.e. 1-65536 may need to be 1-20000, the 20001 to 40000, etc. ! - Analysis may be done on multiple computers to speed solution !******************************************************************** !******************************************************************** ! Instructions: ! 1. Run the sensitivity analysis (tstiff_eigen_analysis.ain) ! 2. Post process the results and determine the 4 most sensitive ! components to torsional stiffness ! 3. Find the four commands "cmsel,s,line#" in this script and change ! "#" to the "line" component numbers to reflect the new results ! 4. Start/resume the new frame database in ANSYS and run the ! create_combo.txt script ! 5. Save the database, exit, and make a copy of it ! ! 6. Modify the /filn command to be /filn,"database name" ! ! 7. Execute in Batch Mode and check parameters.txt every so ! often !******************************************************************** /batch, /filn,07_Optimization resu *DIM,TSTIF,ARRAY,65536,1,1 *DIM,RATIO,ARRAY,65536,1,1 *DIM,ITERWEIGHT,ARRAY,65536,1,1 *DIM,WEIGHT,ARRAY,16,1,1 *DIM,XAREA,ARRAY,16,1,1 dens=0.284 *Get,x1,secp,1,prop,area *Get,x2,secp,2,prop,area *Get,x3,secp,3,prop,area *Get,x4,secp,4,prop,area *Get,x5,secp,5,prop,area *Get,x6,secp,6,prop,area *Get,x7,secp,7,prop,area *Get,x8,secp,8,prop,area *Get,x9,secp,9,prop,area *Get,x10,secp,10,prop,area *Get,x11,secp,11,prop,area *Get,x12,secp,12,prop,area *Get,x13,secp,13,prop,area *Get,x14,secp,14,prop,area *Get,x15,secp,15,prop,area *Get,x16,secp,16,prop,area
! This is the distance in ft between the two forces couple_length=17/12 csys,200 nsel,s,loc,x,0,0.1 *get,d1,node,0,num,min csys,201 nsel,s,loc,x,0,0.1 *get,d2,node,0,num,min csys,202 nsel,s,loc,x,0,0.1 *get,d3,node,0,num,min csys,203 nsel,s,loc,x,0,0.1 *get,d4,node,0,num,min fval=100 csys,204 nsel,s,loc,x,0,0.1 *get,fc1,node,0,num,min csys,205 nsel,s,loc,x,0,0.1 *get,fc2,node,0,num,min ! Enter in the node where vertical displacement should be taken from dnode=fc1 ! Apply Loading condition ! Displacement constraints /prep7 allsel d,d1,all d,d2,all d,d3,all d,d4,all ! Force f,fc1,fz,-fval f,fc2,fz,fval /solu antype,0 allsel solve /post1 set,1 csys,0 *afun,deg *get,dnode_u,node,dnode,u,z *get,dnode_y_loc,node,fc2,loc,y theta=atan(dnode_u/dnode_y_loc) tstiff=((couple_length*fval)/theta) TSTIF(i,1)=tstiff RATIO(i,1)=tstiff/sumweight /prep7 *get,endtime,active,0,time,cpu PARSAV,ALL,'parameters','txt',' ' *ENDDO fini /exit,nosave
Page 3
70
TSMM Post-Processing Script (ctorsion.m) function[avgtstiff]=ctorsion(data) % ctorsion.m % Thomas Steed % University of Cincinnati % E-MAIL: [email protected] % TEL: (513) 260-8955 %************************************************************************** % Description: Calculates the Experimental Torsional Stiffness of a FSAE % chassis for data acquired on the Tosional Stiffness Measuring Machine %************************************************************************** %~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ % Instructions: % 1. Rename the time_data_sum variable from the TSMM *.mat to a variable % called data % 2. Issue the command "ctorsion(data)" at the Matlab prompts %~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ %-------------------------------------------------------------------------- % Step 1: Resample data to smaller working size %-------------------------------------------------------------------------- ch1=data(1:500:length(data),1); % Shaker LVDT 1 ch2=data(1:500:length(data),2); % Shaker LVDT 2 ch3=data(1:500:length(data),3); % Shaker LVDT 3 ch4=data(1:500:length(data),4); % Shaker LVDT 4 ch9=data(1:500:length(data),9); % Strain Gage Force ch10=data(1:500:length(data),10); % Strain Gage Force %-------------------------------------------------------------------------- % Step 2: % Calculate the difference between the next point and the point prior to % locate where the next data step is %-------------------------------------------------------------------------- for i=1:length(ch1)-1 j=i+1; diff1(i,1)=ch1(i)-ch1(j); diff2(i,1)=ch2(i)-ch2(j); end %-------------------------------------------------------------------------- % Step 3: Find the index where the step occurs and determine wether the % index is a point where the load is being applied or relaxed %-------------------------------------------------------------------------- % Find the index where the shaker movement occurs indx1=find((diff1)>=0.010); indx2=find((diff2)>=0.010); indx3=find((diff1)<=-0.010); indx4=find((diff2)<=-0.010); if length(indx1)<length(indx3) indx1=indx3; end if length(indx2)<length(indx4) indx2=indx4; end %-------------------------------------------------------------------------- % Step 4: Calculate the average for each data step %-------------------------------------------------------------------------- ch1_avg(1)=sum(ch1(1:indx1(1))/length(1:indx1(1))); ch2_avg(1)=sum(ch2(1:indx1(1))/length(1:indx1(1))); ch3_avg(1)=sum(ch3(1:indx1(1))/length(1:indx1(1))); ch4_avg(1)=sum(ch4(1:indx1(1))/length(1:indx1(1))); ch9_avg(1)=sum(ch9(1:indx1(1))/length(1:indx1(1))); ch10_avg(1)=sum(ch10(1:indx1(1))/length(1:indx1(1))); for ii=1:length(indx2)-1 ii=ii+1; ch1_avg(ii)=sum(ch1(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); ch2_avg(ii)=sum(ch2(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); ch3_avg(ii)=sum(ch3(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); ch4_avg(ii)=sum(ch4(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); ch9_avg(ii)=sum(ch9(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); ch10_avg(ii)=sum(ch10(indx2(ii-1)+1:indx1(ii)))/length(indx2(ii-1)+1:indx1(ii)); end ch9_avg(ii)=sum(ch9(indx2(ii-1)+1:indx2(ii)))/length(indx2(ii-1)+1:indx2(ii)); ch10_avg(ii)=sum(ch10(indx2(ii-1)+1:indx2(ii)))/length(indx2(ii-1)+1:indx2(ii)); %-------------------------------------------------------------------------- % Step 5: Calculate Torsional Stiffness for each Step %-------------------------------------------------------------------------- twidth=4; % ft for i=1:length(indx2)-1
Page 1
% Average the two force data streams avg(i)=((abs(ch9_avg(i+1)-ch9_avg(1))+(abs(ch10_avg(i+1)-ch10_avg(1)))))/2; % Calculate the degree of twist deg1(i)=atan((abs(ch1_avg(i+1)-ch1_avg(1)))/24)*(180/pi); deg2(i)=atan((abs(ch3_avg(i+1)-ch3_avg(1)))/24)*(180/pi); deg(i)=deg1(i)-deg2(i); % Calculate the numerator to the tstiff equation num(i)=avg(i)*twidth; % Caluclate the torsional stiffness tstiff(i)=num(i)/deg(i); end %-------------------------------------------------------------------------- % Step 6: Calculate the average Torsional Stiffness from all data steps %-------------------------------------------------------------------------- avgtstiff=sum(tstiff)/length(tstiff);
Table of Contents 1. Test Information.......................................................................................................... 3 2. Testing Equipment List ............................................................................................... 3 3. TSMM Fixture Parts List ............................................................................................ 4 4. Pre-Test Notes ............................................................................................................. 6
Table 12: Testing Equipment List Item Description Location Quantity
1 VXI Mainframe Vibes Lab 1 2 Vishay Strain Gage Conditioner and Amp. System Vibes Lab/High Bay 1 3 Dymax 24 Volt Power Supply Vibes Lab Cabinet 1 4 LVDTs Frame Cabinet 3+ 5 Dial Indicators SMLab/SAE Cabinet 5+ 6 MTS 4 Post Simulation System High Bay 1 7 Data Acquisition PC Vibes Lab 1 8 Fire Wire Cable Vibes Lab 1 9 HP ICP Break Out Box Vibe Lab Cabinet 2 10 LVDT Uni-Strut Fixture High Bay 1
11 C-Clamps FSAE & Vibes Lab Toolboxes 3+
12 Safety Straps High Bay Cabinet 6 13 Red Cherry Pickers High Bay 2 14 Rolling Cart FSAE Area 1 15 Weight Set [5-20,1-10,1-5,(Qty-lb)] High Bay 7 16 Video Camera Your House? 1 17 Long BNC Cables (Under Lattice) High Bay 4
G.4
3. TSMM Fixture Parts List
Table 13: TSMM Fixture Parts List
Item Part Description Material/Grade Location Quantity
4. Pre-Test Notes In preparation for testing, organizing equipment and hardware will save time and frustration. It is recommended that the equipment list and procedure are thoroughly read over. Depending upon the extravagance of the suspension some major modifications to the adjustable solid shocks and alignment spacers may be required. Consulting the current suspension team leader for the length and the proper spacer thickness is strongly suggested when using the solid shocks. Any miscommunication between the test conductor and the suspension and/or frame team leader could potentially be very hazardous to the car. Having this person on hand at the time of the test is suggested. However, with the knowledge in this procedure and use of the step by step instructions will create an environment that is safe for all team members and the car. 4.1 “BUY OFFS” This procedure has been written in a manner in which each step is intended to be bought off by the test conductor. It is EXTREMELY important that each step is read and executed as written. If at any time a step is read and not understood testing should halt and a group consensus should be met before proceeding. This is not a test that should be conducted at a time where personnel is fatigued. Each step MUST be bought off for completion to protect the hardware, NO MATTER how juvenile or tedious it might be. Buying off shall be completed by any member in the testing team. “Buy offs” shall be completed by initialing and dating the “Buy off” cell in each step of the procedure. The team member who “Buys Off” takes the responsibility that the step has been completed. 4.2 Bagging & Labeling Throughout the testing process hardware should be bagged and labeled accordingly. Bagging and labeling all the removed parts will prevent any problems remembering the hardware configuration during re-installation. Having extra hardware after re-installing all the parts is not acceptable. Tires should be labeled RF, LF, RR, or LR before they are removed from the car. Each shock should be bagged and labeled with RF, LF, RR, or LR and an orientation arrow to ensure proper re-installation. 4.3 Testing Duration Following this procedure should limit the test duration to no more than 6 hours. However, assuming that Murphy’s Law is always in effect, expect to dedicate a full day to testing. It is absolutely necessary to remove the car from the TSMM fixture at the end of the day. With that said, do not start testing in the late afternoon. 4.4 *****************************E-Stop******************************* The MTS 4 Post Simulation System is equipped with two E-Stop buttons. These buttons are an EXTREME Hazard to the car. If at any time the car is connected to the fixture with solid shocks installed these buttons MUST NOT be tripped. The E-Stop buttons when depressed will cause each post to return to a natural zero based on the hydraulic pressure in the system. These positions are reached very quickly and at different rates. If an E-Stop is tripped there will be damage to the chassis. All members of the test team must be made aware of this.
5.1 In the high bay at the 4 post shaker, remove the center cover of each wheel pan by removing the M4 X 0.70MM X 20MM socket head cap screws (Qty 4). Store the center covers and socket head cap screws in a labeled plastic bag and place in the control room.
5.2 Install the base plate of the TSMM fixture on each post using M12 X 1.75 MM X 120 MM (Qty 4) socket head cap screws. Each screw shall be torqued in a star pattern. **************************NOTE************************ The base plate must be seated against the horizontal surface of each wheel pan to be installed correctly. It may be necessary to lubricate the shoulder of each socket head cap screw to aid in installation. If lubrication is necessary make sure base plate identification labels are not destroyed by the lubricant. *******************************************************
5.3 Install the corresponding labeled aluminum load cylinder on each base plate. Insert a 1/8” Ø brass fuse to orient and connect the base plate and cylinder together.
5.4 Plug in and turn on the strain gauge amplifier. Allow the amplifiers to warm up for 15-30 minutes so that a stable measurement can be made.
5.5 Using the FlexTest software on the pc in the control room turn on the hydraulic shaker system. Make sure that “External” and “High” are selected and press “Run.” Refer to Figure 57 below.
Figure 57: 4 Post Shaker Software Startup
G.8
5.6 Plug in and secure two HP Patch Boxes to the VXI mainframe making sure to install each bank of channels to the corresponding pin connector. (1-4, 5-8 markings)
5.7 Plug in the fire wire cable into the VXI mainframe and then to the back of the PC. Power up the VXI first followed by the PC.
5.8 Start Matlab. Make a new working directory, locate and copy the following scripts: ttest.m, vxiarnge.m, vxierror.m, vxiacquire.m, vxiacquire2.m, vxiinit.m, vxisetup.m, vxi_sae.m
5.9 Unwind the cables for each Load Cylinder and the banana clip connectors into the Vishay 2100 Strain Gage Amplifier system using Figure 58 and Table 14 as a guide. Match the shaker LVDT order sequentially.
Figure 58: Vishay Amp. Banana Clip Plug-ins
Table 14: Strain Gage Wiring Scheme
Wire Color Plug-in Color RED RED
BLACK BLACK SILVER (braided) GREEN
GRAY WHITE Using the BNC to banana clip wires connect the RED wire to the OUTPUT RED plug-in and the BLACK wire to the OUTPUT BLACK plug-in. Using Table 10 as a guide plug in the BNC connectors into the HP Break Out Box.
G.9
5.10 Place the stack of weights outlined in Table 1 next to one of the posts.
5.11 Start the torsion test MATLAB script by issuing “ttest” at the MATLAB command line while in the correct working directory.
5.12 Turn the monitor towards you so that you can see the display as weights are stacked onto the Load Cylinder.
5.13 Confirm that the amount of weight that is stacked onto the Load Cylinder is the value that is displayed on the digital readout on the screen.
5.14 If the displayed value matches the amount of weight that is on the Load Cylinder skip this step and step 5.16. If the displayed value does not correspond with the amount of weight that is stacked on the Load Cylinder, use the values on the Strain Gage Amplifier display to create a new calibration curve. Plot this curve, take its slope and enter in the new calibration value into the top of the “ttest.m” MATLAB script.
5.15 Repeat steps 5.11 to 5.15 for the remaining Load Cylinders.
5.16 Unplug and coil the Load Cylinder cables from the Vishay Strain Gage Amps and place them with each post so that they can move together. **********************WARNING*********************** Make sure to support each Load Cylinder cables so that their weight is not pulling on the strain gauge. *******************************************************
5.17 Turn on the green air compressor in the high bay.
G.10
5.18 Measure the length of each shock as the car supports it’s own weight and record in Table 15:
Table 15: Ride Height Shock Lengths Location Length Left Front
Right Front Left Rear
Right Rear
5.19 *******************Multiple Person Operation*************** Lift the car onto a rolling cart. Be sure that the weight of the car rests on the frame and not the suspension components. Lift the car by having a person at each corner bear hug the wheel.
5.20 Remove each wheel, label and set aside. Make sure to place each wheel lug in a safe and easily accessible place for use in the next step.
5.21 Install a TSMM fixture Hub Adapter (Qty 4) on each corner of the car, using the car’s wheel lugs. Hand tighten each nut until it is seated at the outer face of each adapter. Using a 19 MM open ended wrench tighten each wheel lug in a star like pattern. It may be necessary to have another person hold the brakes while the lugs are tightened.
5.22 Return to the control room and iterate on the setpoint of each post until the External Readout reads the same for each post. Type in the setpoint box, hit enter, and watch the external readout value. Repeat until all the Shaker LVDTs read the same.
Figure 59: 4 Post Height Software Commands
G.11
5.23 Consult the current suspension team leader if doubtful of measurement accuracy of Table 15. Adjust the length of the solid shocks to match that of Table 15 and label accordingly.
5.24 *******************Two Person Operation****************** While supporting the suspension at the hub, remove the shock/damper by backing out the ¼” – 28 bolts at the bell-crank and frame mount. Make sure to carefully watch for any loose spacers. Bag, label and store shock and any necessary hardware.
5.25 Install solid shock with alignment spacers using ¼”-28 bolts. Properly torque each bolt.
5.26 Repeat steps 5.25 and 5.26 for each corner of the car.
5.27 Raise the blue lift in the high bay all the way up and rotate the support arms away from the 4 Post Shaker area.
5.28 Roll the cart/car into the middle of the 4 Post Shaker area so that the front point towards SMLAB.
5.29 Place a cherry picker at the front of the car. Install a strap on the cherry picker’s hook/clip and raise it so that the boom will not hit the car. Position the cherry picker so that the strap wraps around the front end and supports the bottom frame rails.
5.30 Roll the second cherry picker to the rear of the car. Install a strap on the clip and raise the boom. Position the cherry picker so that the strap will wrap around the jack bar.
5.31 With another person raise each cherry picker together so that the car will clear each post.
G.12
5.32 Lower the blue lift a small amount and as a safety measure wrap a strap around the bottom side impact tube of the frame and up to each arm. Provide each strap with ample slack so that when the frame is attached to the fixture it will not be supported.
5.33 Return to the control room and check that the external readout of each post’s LVDT still reads as it did before. If it has made any significant change more time is needed for the fluid to reach a constant operating temperature. Wait 10 minutes if LVDTs are still fluctuating.
5.34 Make any adjustments to the height of the rear of the car so that the hub adapter and load cylinder pivot points match.
5.35 **********************WARNING*********************** Steps 5.36 through 5.47 must be completed in the sequence described. If not the potential of pre-loading the suspension/frame is greatly increased. *******************************************************
5.36 Install the air supply at the right rear post and turn on. Gently float the post into a planar position so that the car can be translated with the cherry picker so that the slot of the Hub Adapter can slide around the Load Cylinder. Turn off the air supply!
5.37 Make any vertical adjustments by raising/lowering the cherry picker so that the pivot holes match in each part. Install the pivot bolt and hand tighten the nut. *************************NOTE************************* A hammer may be necessary to TAP the pivot bolt in. *******************************************************
5.38 Remove the air supply to the right rear post and install on the left rear post. Turn the air supply on.
5.39 Wrestle the left rear post so that it is floating and can be easily moved with a small amount of force. Make sure that the post is floating and not oscillating because of low air pressure! Very gently float the post so that the Load Cylinder is in the slot of the Hub Adapter. Turn off the air supply!
G.13
5.40 Make any out of plane adjustments by raising/lowering the cherry picker so that the pivot holes match in each part. Install the pivot bolt and hand tighten the nut. **************************NOTE************************ The air supply may be turned on to make small adjustments if necessary. *******************************************************
5.41 Adjust the height of the front of the car by raising/lowering the boom of the cherry picker. Position the height so that the pivot point of the Load cylinder matches with the Hub Adapter.
5.42 Remove the air supply and install on the front left post. If necessary wait for air pressure to return before turning on.
5.43 Wrestle the left front post so that it is floating and can be easily moved with a small amount of force. Make sure that the post is floating and not oscillating because of low air pressure! Very gently float the post so that the Load Cylinder is in the slot of the Hub Adapter. Turn off the air supply!
5.44 Make any out of plane adjustments by raising/lowering the cherry picker so that the pivot holes match in each part. Install the pivot bolt and hand tighten the nut. **************************Note************************* The air supply may be turned on to make small adjustments if necessary. *******************************************************
5.45 Remove the air supply from the left front post and install on the right front post. Wait until air pressure has been completely been restored.
5.46 Wrestle the right front post so that it is floating and can be easily moved with a small amount of force. Make sure that the post is floating and not oscillating because of low air pressure! Very gently float the post so that the Load Cylinder is in the slot of the Hub Adapter. Turn off the air supply!
5.47 Make any out of plane adjustments by raising/lowering the cherry picker so that the pivot holes match in each part. Install the pivot bolt and hand tighten the nut. **************************Note************************* The air supply may be turned on to make small adjustments if necessary. *******************************************************
G.14
5.48 Lower and roll cherry pickers away from testing area.
5.49 Roll the cart out from underneath the car.
5.50 Position LVDT Fixture underneath the car and place some weight on the lower rails to hold it in position.
5.51 Plug the Shaker LVDTs into Channels 1-4. **************************Note************************* These cables can be found lying along the control room wall. *******************************************************
5.52 Unwind the cables for each Load Cylinder and the banana clip connectors into the Vishay 2100 Strain Gage Amplifier system using Figure 60 and Table 16 & 10 as a guide. Match the shaker LVDT order sequentially.
Figure 60: Vishay Amp. Banana Clip Plug-ins
Table 16: Strain Gage Wiring Scheme
Wire Color Plug-in Color RED RED
BLACK BLACK SILVER (braided) GREEN
GRAY WHITE
5.53 Attach the three Frame LVDTs to the fixture using the C-Clamps and record their location in Table 17.
G.15
5.54 Connect the BNC attached to each Frame LVDTs to the HP Patch Boxes according to Table 21.
5.55 Connect the Frame LVDTs 24 VDC and ground wires to the DYMAC power supply. The black wire is 24 VDC and the silver wire is the ground. Refer to the Figure 61 below.
Figure 61: DYMAC Power Supply Terminals
**************************Note************************* Only one power supply is needed to power all three LVDTs. Install two wires in the first two terminals. *******************************************************
5.56 Plug in the DYMAC 24 Power Supply.
5.57 Start the torsion test MATLAB script by issuing “ttest” at the command line and manually operate each Frame LVDT to confirm that it is operating within the range of each plot window.
5.58 Attach the dial indicators to the LVDT Fixture using a magnetic post and record their location in Table 17. **************************Note************************* It is much easier to log displacement data from the dial indicators if two people record what is happening from each side of the car at each displacement point of the 4 post shaker. *******************************************************
5.59 Print out an extra set of Tables 7 & 8 for recording displacements from the dial indicators.
G.16
5.60 Zero the strain gage amps by turning the “Balance” knob so that the red lights are not on and the display reads very close to zero if not zero.
5.61 Start the video camera recording.
5.62 Start the MATLAB torsion test script by issuing “ttest” at the command line.
5.63 Record the external readout of SHAKER LVDT 25 & 26 below: SHAKER LVDT 25 External Position:_________________ SHAKER LVDT 26 External Position:_________________
5.64 With the test conductor in the control room, begin the first test as outlined in Table 18 by moving the setpoint in 0.020” increments. Record the deflection from the dial indicators at every increment position in Table 18.
5.65 Save the MATLAB workspace to a file with the formatted name: “TSMM_CARNAME_DATE_20_MIL.mat.”
5.66 Save the MATLAB display with the formatted name: “TSMM_CARNAME_DATE_20_MIL.fig.”
5.67 Unload the frame in four 0.050” increments until the neutral positions recorded in step 5.61 are achieved.
G.17
5.68 Pause the video recorder.
5.69 Re-zero the dial indicators.
5.70 Re-zero the strain gage amps.
5.71 Start the video camera recording.
5.72 Start the MATLAB torsion test script by issuing “ttest” at the command line.
5.73 Record the external readout of SHAKER LVDT 25 & 26 below: SHAKER LVDT 25 External Position:_________________ SHAKER LVDT 26 External Position:_________________
5.74 With the test conductor in the control room, begin the second test as outline in Table 19. Record the deflection from the dial indicators at every increment position in Table 19.
5.75 Save the MATLAB workspace to a file with the formatted name: “TSMM_CARNAME_DATE_50_MIL.mat.”
5.76 Save the MATLAB display with the formatted name: “TSMM_CARNAME_DATE_50_MIL.fig.”
G.18
5.77 Pause the video recorder
5.78 Unload the frame in four 0.050” increments until the neutral positions recorded in step 5.73 are achieved.
5.79 Make a backup copy of the data files and figures onto a removable media storage device.
5.80 Roll the two red cherry pickers back into their previous positions, install the straps and jack the cherry pickers up until the straps are taught.
5.81 Remove the pivot bolt from each Load Cylinder.
5.82 ******************Three Person Operation****************** With a person in between two shaker posts jack up each cherry picker simultaneously until the Hub Adapters are clear of all the Load Cylinders. **************************Note************************* The person in the middle of the shakers should have a hand on the frame to stabilize it once the Hub Adapters clear the Load Cylinders.
5.83 Turn off the hydraulic pressure to the shakers by using the software in the control room
5.84 Unplug and detach all instrumentation that is attached to the LVDT fixture. Coil and tape all instrumentation wires and store the instrumentation in the frame cabinet.
G.19
5.85 Remove the LVDT fixture from underneath the car and place it along side the railing in the high bay.
5.86 Unplug and coil the Load Cylinder cables from the Vishay Strain Gage Amps and place them with each post so they can move together. **********************WARNING*********************** Make sure to support each wire so that its weight is not pulling on the strain gage. *******************************************************
5.87 Attach the air supply to each shaker post and float the post out of the way to make room for the rolling cart.
5.88 Roll the cart underneath the car.
5.89 Remove the security straps from the frame to the lift arms and store the straps in the high bay cabinet.
5.90 Lower the car onto the rolling cart, detach the straps on the cherry picker and store the cherry pickers where they were located previously.
5.91 Roll the car/cart out from the testing area.
5.92 Remove each solid shock assembly from the suspension system and replace with the bagged hardware.
G.20
5.93 Remove and properly store all TSMM fixturing and hardware back in the frame cabinet.
5.94 Return the VXI mainframe, patch boxes, DYMAC power supply, and PC to the vibrations lab.
5.95 Turn off the Vishay 2100 Strain Gage Amplifier System.
5.96 Install the center cover of each wheel pan back onto each post.
G.21
6. Appendix A: Sensor Record Sheet
Table 17: Transducer Log
Figure 62: TSMM Sensor Locations
Location Sensor Description Model Number Serial Number Calibration 1
2
3
4
5
6
7
8
9
10
G.22
7. Appendix B: TSMM Test Data Record Tables
Table 18: TSMM Test 1 0.020" Increment Location # (Refer to Figure 62 CAR NAME:
Defl. 1 2 3 4 5 6 7 8 9 10 0.020
0.040
0.060
0.080
0.100
0.120
0.140
0.160
0.180
0.200
Table 19: TSMM Test 2 0.050" Increment Location # (Refer to Figure 62) CAR NAME:
Defl. 1 2 3 4 5 6 7 8 9 10 0.050
0.100
0.150
0.200
G.23
8. Appendix C: Equipment Record
Table 20: Equipment Record
9. Appendix D: Channel Record
Table 21: Channel Record Channel # Description Location
1 Shaker LVDT
2 Shaker LVDT
3 Shaker LVDT
4 Shaker LVDT
5 Strain Gauge
6 Strain Gauge
7 Strain Gauge
8 Strain Gauge
9 Frame LVDT
10 Frame LVDT
11 Frame LVDT
12 Frame LVDT
Item Description Model Number Serial Number 1 PC Station 2 HP Breakout Box 3 HP Breakout Box 4 Dymax Power Supply 5 VXI Main Frame 6 Strain Gage Amp 7 Strain Gage Amp 8 Strain Gage Amp