Hyperworks Trail vs Educational / Ship - Windturbine Collision Analisus
Altair EmployeeI have an urgent question right now,
I do have access to a trail licence for my research, however I could not setup the licence server, hence me using the educational version on my laptop.
I forgot about the 100k node limit, and basicaly have left my laptop meshing for 7 days on a ship hull. Now the student licence has the Radios Program build in,
Is it at all possible to run the radios simulation without saving in the student licence? or did I thanks to my mistake just lose 7 days?
Seccond question:
First I wil go into the background of the casus
I am doing my graduate research as Student Naval Architecture for the Dutch Coastguard. I am looking into the effect or risk the current and upcomming wind turbine farms on the North Sea create regarding Ships 'Not Under Command', which means they either lost propulsion or steering or both, which means they are drifting where the wind and tidal current brings them. Next to risk areas and the impact analysis I also look at how to lower the risks, and placement of one or multiple emergency towing vessels.
Within Hyperworks, and as I have been adviced by some of the people from Altair, I am prepping for Radioss,
Basically I have 20 meter long halfcut sections of different ships, next to that I have a wind turbine foundation as commonly used on the north sea,
Both objects have been made midsurface meshes off, the foundation took 7 days to mesh, one of the bigger shiphulls with 18 days continuous is still meshing.
Both objects will get basic steel properties E=210000 and poison of 0,26
The shiphulls will get a mass equall to the mass of the entire loaded ship, though as of this moment I am not sure whether to apply this mass as force or list is as mass of the unit.
The turbine foundation will be clamped solid at the seabottom instead of beeing clamped in with dampened springs (soil) to make the calculation simpeler.
The drag and other water effects are neglected for simplification
Both objects have to deform and desirably the shiphull will have to tear.
Desired outcome:
The basic point of the above is to figure out at what speed at the point of impact the shiphull will tear,
As far as I have understood it now I will have to run the simulation multiple times to figure out at what speed this will occur.
Why:
When the shiphull tears the cargo contents will spill or worse, the ship will sink.
This will lead to an environmental disaster, that can be calculated back to environmental damage and human casualties based on existing examples in the past.
I have real data of drifting speeds over the past 5 years, I will be able to compare the speed with real life data, and then can narrow down if a ship basicaly drifting by wind and tidal current can split open by colliding with a windmill and at what wind speed, in beafort, that will likely occur.
The questions sumerised:
Hyperworks student edition:
- Is it at all possible to run the radios simulation without saving in the student licence? Or did I thanks to my mistake just lose 7 days?
Hyperworks 2017 Trail:
- The shiphulls will get a mass equall to the mass of the entire loaded ship, though as of this moment I am not sure whether to apply this mass as force or list is as mass of the unit. How woud you advise to apply this mass
- As far as I have understood it now I will have to run the simulation multiple times to figure out at what speed this will occur. Is there another method?
Thank you very much,
A J Prins
Answers
-
Hi,
You cannot run the simulation in Student Edition due to the node restriction. If you have a full license you can save from Student Edition and can run the same in full license.
I would recommend you to try to model this load in the ship as mass. You can apply the same using ADMAS.
There is no other go, you may have to run several iterations to find the speed.
0 -
Thank you George,
I will uninstall the student edition from my laptop, fix the licence server and get the Trail licence installed on the laptop aswell
I will update the post with news if I have some,
Given that meshing takes 7- 20 days how long would you expect the simulation to run?
0 -
Hi,
Obviously it will take more time to run and it depends on the time step of the model.
Even in some cases simple models will take more time to run. Once the model set up is done we will check the time step of the run and if it's very less we will try to control the time step and thereby we can try to speed up the run.
0 -
So one of the two hulls is now done meshing, though something odd happened.
It put the hull together in one part with the Windturbine foundation,
I am now figuring out if I can split the two
0 -
<?xml version="1.0" encoding="UTF-8"?>
When I select the mesh it selects them both
And there is for sure a gap here... Both objects are on Middle surface now...Edit:
I figured out where I went wrong,I sorted the midsurface to the wrong component,
Is there any way to correct my mistake?
Since I have the windturbine as a seperate element saved, is deleting the foundation only an option?0 -
Hi,
You can organize the elements in their respective collector.
Select the element and use 'by attached' option so that you may be able to select these elements and can organize in other collector.
0 -
Somehow when I do that it keeps selecting both of them...
Also with how big the models are, every step I try locks me out of the program for 3 minutes.
0 -
Hi,
If you can share the model file through the secure dropbox we can also try from our end.
0 -
Its uploading, it will take some time...
0 -
The file should be uploaded,
Thanks so much0 -
I made a small mistake, it is in the link in the message, however I will upload it to the secure dropbox aswell,
Due to its size it will need 1:40 to upload
0 -
Please use the link in the message, the transfer estimates 20 days
0 -
I think the secure dropbox transfer also has finnished,
Did you have the oportunity to take a look?
0 -
Hi,
Reviewed your model and has sent the modified file.
0 -
Thank you,
I just have been able to set up the boundry conditions and added the mass,
Now I am setting up the radioss simulation0 -
Hmm, I made some mistakes somewhere,
I'm backtracing the errors, will report my findings:.. BLOCK FREE FORMAT
.. STARTER RUNNING ON 8 THREADS
.. CONTROL VARIABLES
.. UNITS SYSTEM
.. MATERIALS
ERROR ID : 683
** INPUT ERROR IN MATERIAL LAW
DESCRIPTION :
-- MATERIAL ID : 1
-- MATERIAL TITLE : Hull Shell
DENSITY IS LOWER THAN OR EQUAL TO ZERO
ERROR ID : 683
** INPUT ERROR IN MATERIAL LAW
DESCRIPTION :
-- MATERIAL ID : 2
-- MATERIAL TITLE : Foundation Shell
DENSITY IS LOWER THAN OR EQUAL TO ZERO
.. NODES
.. PROPERTIES
ERROR ID : 178
** ERROR IN PART DEFINITION (PROPERTY)
DESCRIPTION :
-- PART ID : 6
-- PART TITLE : component1
PROPERTY ID=0 DOES NOT EXIST
ERROR ID : 179
** ERROR IN PART DEFINITION (MATERIAL)
DESCRIPTION :
-- PART ID : 6
-- PART TITLE : component1
MATERIAL ID=0 DOES NOT EXIST
.. 3D SHELL ELEMENTS
** ERROR TRAPPED : REPORT TO LISTING FILE
ERROR ID : 760==== End of solver screen output ====
==== Job completed ====0 -
RIght,
I think I have fixed most,
Just one thing I do not understand so far,
The admas menu, it comes from view -> utility's -> AdmasFor this one I selected parts, components, and my component, I also used the seccond option as the mass is added to the entire component and not all the nodes.
Then it needs a visualisation node,
I just made one for it guessing it will work, what is it exaclty I need to pick for that?0 -
Hi,
The above error relates to the material created, please check the material values. Hope it is sorted out.
Select only one node as visualization node as this is just only for the visualization purpose.
0 -
I got a lot further now:
I am doing a model check next:.. BLOCK FREE FORMAT
WARNING ID : 1114
** WARNING WHILE READING
DESCRIPTION :
THERE IS NO DEFINED PART IN INPUT FILE... STARTER RUNNING ON 8 THREADS
.. CONTROL VARIABLES
.. UNITS SYSTEM
.. MATERIALS
.. NODES
.. PROPERTIES
.. 3D SHELL ELEMENTS
** ERROR TRAPPED : REPORT TO LISTING FILE
ERROR ID : 760==== End of solver screen output ====
==== Job completed ====
0 -
well this is quite a lot of errors to go through:
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi,
Can you please share the starter out (_0000.out) file, which will list the modelling errors.
0 -
File send,
The difference over the earlier send files are as follows:
The shiphull and turbine have had their geometry's swapped so they are with their meshes.
The shiphull gets via admass the full mass of the ship and its cargo, in case of the smaller vessel this is 7500 metric ton or 7500000 kg
Furthermore the hull gets a velocity of 3.5 kts or 1.800556m/s (which is for the smaller vessel a possible driftspeed at 6 on Beaufort)The foundation gets fully clamped without dampening at its bottom.
the material for both objects has been determined as basic steel with the following properties:
Density: 7860kg/m3
Emodulus: 2100000
Poison ratio: 0.26
-edit-
I forgot,
Timebase for Radios was set at 100
and the rest of the frequency's at 0.05
These values where added in the engine starter assistant
and well, that apperantly delivers the above result
0 -
Hi,
In the out file it's not specifically showing the error, but ensure that PART card image is provided to the components and confirm material/property are also assigned.
0 -
I have fixed the fact that the part card image was not provided, and still the same error shows.
Also my biggest shiphull finished meshing, I started to fill in the boundry conditions there, and as soon as I put in the initial velocity and hit create, the program freezes. It has been frozen now for 5 days straight, is it safe to asssume the program is bricked at that point?
I decided to try the workflow with a small model to try. while the tutorials worked I wanted to try it out as I do with the shiphulls
So I did the same steps with a small model. I made a round shaped solid object and a small plate in rhinoceros. I exported the geometry in Iges and imported this in Hypermesh. I started the midsurface mesh, and after the mesh was finished, I assign the part cards, the material values:
Density: 7860kg/m3
Emodulus: 2100000
Poison ratio: 0.26
The rest I leave default
and the property card in which I do not change anything from defaultThen I move over to the ADMAS tool and assign a mass to the plate with adv01??
in this case 700kg
I add a initial velocity of -1m/s to the plate in X
I add the boundary condition that the round pice is clamped in the bottom with a few of the bottom nodes.I setup the engine file manager with the following info:
Timebase for Radios was set at 100
and the rest of the frequency's at 0.05
I still get the same error... just as with the small hull I sended over the dropbox, the big hull I have not even tried yet as well it is stuck with the initial velocity setup...
Am I doing something colosaly wrong? Is there some other tutorial I need to take a look at to look at the things I need to setup that I am not aware of?I have uploaded and attached the small try, in case anyone wants to take a look.
Thanks in advance
A J Prins0 -
Hi,
The issue with this model file is you have assigned shell property to solid component. For the solid component (Round) assign P14_Solid property and run the analysis.
0 -
This fixed the issue,
that makes me wonder if I should assign a different property to the bigger shiphull and the turbine foundation. They are both Midsurface Meshes and have been assigned P1 SHELLMeanwhile I noticed something odd with the results:
<?xml version="1.0" encoding="UTF-8"?>
ANIMATION FILE: try4A387 WRITTEN RESTART FILES: try4_0001_[0001-0008].rst WRITTEN ** CPU USER TIME ** #PROC CONT.SORT CONT. F ELEMENT KIN.COND. INTEGR. I/O TASK0 ASSEMB. RESOL 1 .0000E+00 .0000E+00 .2500E+00 .1562E+00 .7812E-01 .5484E+01 .6562E+00 .7812E-01 .6750E+01 2 .0000E+00 .0000E+00 .2500E+00 .1719E+00 .1094E+00 .5625E+01 .1047E+01 .9375E-01 .9406E+01 3 .0000E+00 .0000E+00 .2031E+00 .1406E+00 .1406E+00 .6109E+01 .1500E+01 .7812E-01 .8922E+01 4 .0000E+00 .0000E+00 .2656E+00 .1094E+00 .1562E+00 .4578E+01 .8594E+00 .7812E-01 .7047E+01 5 .0000E+00 .0000E+00 .1719E+00 .1094E+00 .1719E+00 .5672E+01 .2016E+01 .1406E+00 .9125E+01 6 .0000E+00 .0000E+00 .1875E+00 .2500E+00 .7812E-01 .5750E+01 .2969E+01 .1406E+00 .1048E+02 7 .0000E+00 .0000E+00 .1875E+00 .1875E+00 .1406E+00 .6625E+01 .2969E+01 .6250E-01 .1117E+02 8 .0000E+00 .0000E+00 .1250E+00 .1562E+00 .1250E+00 .6438E+01 .2109E+01 .6250E-01 .1105E+02 ** SPMD COMM. TIME ** #PROC FORCES RBY.FOR. RBY.VEL. VELOCITIES TOTAL % CPU 1 .0000E+00 .0000E+00 .0000E+00 .0000E+00 .0000E+00 0.00 2 .1469E+01 .0000E+00 .0000E+00 .0000E+00 .1469E+01 15.61 3 .2969E+00 .0000E+00 .0000E+00 .0000E+00 .2969E+00 3.33 4 .6250E+00 .0000E+00 .0000E+00 .0000E+00 .6250E+00 8.87 5 .7812E-01 .0000E+00 .0000E+00 .0000E+00 .7812E-01 0.86 6 .3594E+00 .0000E+00 .0000E+00 .0000E+00 .3594E+00 3.43 7 .2188E+00 .0000E+00 .0000E+00 .0000E+00 .2188E+00 1.96 8 .1172E+01 .0000E+00 .0000E+00 .0000E+00 .1172E+01 10.61 ** CUMULATIVE CPU TIME SUMMARY ** CONTACT SORTING............: .0000E+00 0.00 % CONTACT FORCES.............: .0000E+00 0.00 % ELEMENT FORCES.............: .1641E+01 2.22 % KINEMATIC COND.............: .1281E+01 1.73 % INTEGRATION................: .1000E+01 1.35 % ASSEMBLING.................: .7344E+00 0.99 % OTHERS (including I/O).....: .6930E+02 93.70 % TOTAL......................: .7395E+02 100.00 % ** MEMORY USAGE STATISTICS ** TOTAL MEMORY USED .........................: 374 MB MAXIMUM MEMORY PER PROCESSOR...............: 52 MB MINIMUM MEMORY PER PROCESSOR...............: 44 MB AVERAGE MEMORY PER PROCESSOR...............: 46 MB ** DISK USAGE STATISTICS ** TOTAL DISK SPACE USED .....................: 57886 KB ANIMATION/TH/OUTP SIZE ....................: 56141 KB RESTART FILE SIZE .........................: 1745 KB ELAPSED TIME = 35.32 s 0:00:35 ESTIMATED SPEEDUP= 2.09 NORMAL TERMINATION TOTAL NUMBER OF CYCLES : 387 License warning One or more of the required features checked out will expire in 22 day(s) ==== End of solver screen output ==== ==== Job completed ====there was no collision between the objects so something went wrong.
(BTW the biggest hull finished assigning the velocity after 5 days, I have no clue what make the program do so long over assigning it...)
0 -
Hi,
You need to define an interface between these two components and then only the contact will be identified. Else it will pass through the other object.
Define a Type 7 interface and you can see the steps for the same at https://www.youtube.com/playlist?list=PLQ7KUGFuVz4u5y3SvIEIU07427-gATaG6 .
0 -
I keep getting these:
MESSAGE ID : 205 ** RUN KILLED : ENERGY ERROR LIMIT REACHED MESSAGE ID : 207 ** RUN KILLED : TOTAL MASS ERROR LIMIT REACHED NC= 98 T= 2.1872E-02 DT= 1.8144E-08 ERR= 99.9% DM/M= 7.7881E+24 ELAPSED TIME= 10.84 s REMAINING TIME= 49558.38 s ANIMATION FILE: try7A002 WRITTENI did follow the steps in the video for the contact card and the output blocks...
I did notice I used the wrong material card, and that I needed to add things in properties
I will try the small shiphull soon
I also setup the wrong speed and density:Due to me working in:
mm/kg/s
I need to work in mm^3/kg and mm/s instead of m^3/kg and m/s...
THose errors have been correctedIshell is now 24 and Ithick is now 1
the material is now m36 plastish tab with a curve.
I also think about only applying the admass on the outermost edges where it would apply on the ship,
Another issue that has not been accounted for is the bending stresses already in the hull due to buoyancy and the ships own weight, I guess, adding those in will be making the simulation even more complex and we would reach a new level of impossible?0 -
Hi,
The mass error (DM/M) in the above out file seems to very high. The acceptable range is 3-5% (.03-.05). Reduce the imposed time step value (Tmin in /DT/NODA/CST).
Check the interfaces for any penetrations.
Ensure that the unit consistency is followed, else it will show very high run time and wrong results.
It seems like it will become more complex but let's see.
0 -
Hmm, even with that, the error keeps showing up,
I have not checked for the penetrations, thats what I am doing next,
<?xml version="1.0" encoding="UTF-8"?>
I did however find issues with the thicknesses, I am worried that the midsurface mesh did not assign the thicknesses, (or did but somehow lost the values)
This would be a disaster in case of the shiphulls as they took over 25 days to complete...0
