IF you do what I suggested you will be ok. You can then run the strayline plugin which is free and get the idea of problem areas and fix. I found 145 problem areas in a quick check and they were repairable. CAD files are notorious for this type of problem. The problem of fixing these types of problems with plugins is they usually assume some type of spacing error and don't always fit you case. Plus if it is an spherical error then geo can changes. I tried ThomThoms clean up and but it was deleting geo that should stay. He also has what is called plan tool but have not used so don't know if that will help.
You just need to hang in there. It will take a little work.

TIG'S tools are great but they will not create the correct surface until you give them good information. Form your PICS that does not look like the case. The reason I gave you the Delft link was to show you what a good model is not that you should use it to make your model. The Bow wrinkle IMHO is caused by inadequate surface definition.
Hope I'm wrong
Still waiting for you to post yor model => that is always the best

Just a quick check I note that your primitive geometry is not drawn on layer zero. SU uses layers for visibility control( layers) only via components. Don't know if this is your problem but check is in order Once I move them back to layer zero I used thoThoms selections toys and moused over some that are not forming and they do not show closed loop so may want to check that also.

Here is a link to one of the supposedly best free hull modelers. http://www.delftship.net/?height=896&width=1680 There site has a number of free models also you can import in 3d dxf. Can do that in 7 but not 8. The PDF they have gives you a good idea about the process.
My guess you need beter splines on the surface generation and is reason for the wrinkles you see. Need to see the model at least the first inside shape.

From the PIC it looks like you have a number of model problems . Is is possible to post you model to 3 d warehouse so we can take a look at in detail and maybe help?

Think you completely misunderstand. It is not uncommon for large( Yes I know yours will not be nearly this large) project to take maybe ten years from start too finish with many up front trade off designs to look at how one wants to accomplish the intended function(s). There will be many desing /schedule / cost studies etc. I think where Charly and I came from is the detail of the hospital vs how it fits in with the functions to be performed. You have targeted lay person and high school students in your original OP. So now what do you want to convey to them in your virtual world and what do you expect them to have when they get done with the simulation or tour? Also this will not be free so you must get funding form maybe some NPO OR ? and will need enough info to get the required help. At present I not not think you can submit coherent request to ask for a grant For example, if this would be used in a school setting or at home then you must consider what the target population has at their disposal to even participate. This leads me to the conclusion the building of the hospital model is the least of your worries and you are possibly( at this time ) dealing with the wrong technical disciplines. I guess I am saying you must first determine what the capability of the target population has at their disposal before you can go off and build something they can use. Based on you location I guess this maybe persons that do not have access to all the bells and whistles others may have. I am guessing you should be talking to some computer / network types => all system design folks? Once you can put on paper a top down functional requirement ( This is not a design requirement ) it comes latter then you can go to them. This will I am sure this will be an iterative process=> the more you learn the better you can refine.
I am even thinking you are in the wrong forum for your effort
Sorry if I am still way off base

Sorry I have to again state form follows function and your previous statements you don't have any budget restrictions is not correct. You have seen AVATAR I guess and they probably spent millions if not hundreds of millions for software, programs , hardware etc. I just cannot rationalize you stated goals on the OP with this obsession with the hospital layout when there are really more pressing problems than it. If you are going to make your project a success then spending time to decided what functions are really needed and a schedule so people asking all the questions can really understand what is required. Is your goal to teach the high school students how to park in a hospital, find there way around one ,or have some type of medical science simulation that they can step through and learn something. Are they off site or in one location and will you be using existing simulations, programming as part of this or just what?
Sorry as you can see you have yet to convey your concept to me
Front to back or back to front has no meaning to me expect your are heading for diaster

@pitrak said:

Another question re: SU and Win 7 64bit:

32bit apps can only use 2GB at a time. But if I have multiple instances of SU open, will they all have their own 2GB of RAM allocated?
I'm planning on buying an i7 quad laptop with 8 or 16 GB of RAM, so would be interested to know this before I buy.

Found this, but is says that the principal does not apply to older software. And as SU is not always state-of-the-art in terms of resource use, I'ld just like to know for sure..
http://www.sevenforums.com/general-discussion/145160-whatis-mem-use-multiple-instances-same-32bit-program-win7-64bit.html

Thanks!

The answer is no. Applications do not use RAM but they use virtual memory and the sum for all applications running can be greater than the RAM. Is is a complex subject see this link for details http://blogs.technet.com/b/markrussinovich/archive/2008/11/17/3155406.aspx

Zoom
Looking at the Intel slash sheet seems to indicate the max single core with turbo boost is 3.7 GHZ I built my system more than 5 years ago and uses a 3.2 GHZ P4 with a 865 PERL mother board. My primary limit is the video =>my board only supports AGP 8x. So pay very close attention to the mother board and the graphics you select. That will probably be the prime bottle neck.
BTW also pay close attentiom to the SKU version of W7 x64 you buy some only handle 2 GB RAM . I would shoot for at least 8GB. I think that is the max you can get now. Servers go to 128GB. Just like the x32 the RAM gets reduced by the the other devices so a large graphics memory will sub from that but then gets remap above the 4GB boundary.

zoom
ome additional info for you:

1. The windows memory manager( W x32) allocates 1/2 of the ram to each of application and the kernel so your 2 GB RAM is low. If you install more RAM then using the option of 4G tuning is possible if your system supports that. It would then allocate memory as 3GB process and 1 GB to kernel. I no not know what the speed implication would be but that allowed me to open a 340MB model but was still slow. That allows you to implement layer control etc to reduce rendering loads . Saving and opening would still be slow however;
2. Check you BIOS and see if turning off hyper threading can help;
3. If you go to w7 x64 then all new drivers are required also. SU will run as a x32 application;
4. My personal preference is to avoid over clocking if possible. One of the main contributors to electronics reliability is emitter base junction temps and that just drives the temps up. Usually we like to keep them down to 125 C or less. I think the upper limit on a single core is about 3.5 GHz and SU does not use dual cores yet. If you have the foot print and connection for more case fan that could be a good move. I have two and still runs warmer that I like;
5. Since it sounds like SU is your secondary use my approach would be to upgrade memory and implement techniques to reduce render load during your modelling. Save and open will still be slow however. Those two should be simple and quick for you to do and then if you are not happy an up grade will be required. Probably lose a few $ on the memory cost. Quad core is the wave of the future so you will have the decision of the apps you run and what they can run to make that.
   See these for more info
   http://sites.google.com/site/sketchupsage/faster and
   http://sites.google.com/site/3dbasecamp2008 and
   http://sites.google.com/site/3dbasecamp2008/all-sessions-2008/working-with-large-models
   Good Luck

one additional comment:
If I scale an entity in a drawing , copy and past into a different target model it comes in at that scale and the units are consistent with the template used in the target drawing. I do have the option in this target drawing to context click on that entity if it was a component to reset the scale without using the tape tool. If it was saved and I then import to the target drawing then that option is not available , but reported units are still consistent at the scaled value.
This is for skp to skp model but form other programs then you may have to use the options avaiable when importing or have some type of truth.

@sonofmoose said:

Hello

"I am a registered medical practitioner with a research Masters degree in Psychiatry and Mental Health who is extremely interested in the ability of computers to play an important role in both professional and lay person healthcare education. As a direct result of this interest, I have been working for the past six years on my proposed freeware healthcare education application (simulation) that is initially aimed at high school students."

Not in healthcare field nor architecture but have done technical management of 1/2 billion dollar programs. As you are well aware form follows function and I see your goal above but suddenly find you in the detail of the structure design, water features, parking grages etc. with a considerable time elapsed and I ask my self the question are you stove piped ? I see no relationship to the following discussions and your goal above. I may be worth your time to set down and draft a spec. or specs ( In this case it would be a software or model requirements ). My contention is your goals can be more quickly obtained by possibility a number of modules vs trying build a very complex initial model. Once those basics are done then the integration of those into a much larger model can be done on a timed sequenced basis?
Sorry if I am completely off the mark but I think your lack of progress, the number of questions and vol. support etc. some what supports my thoughts.

SoM [Jeremy]

From your specs and model size you should not have memory issue unless you are using grahics vs what you noted with large memory or maybe dual monitors. Their memory gets subtraced from what you can use because w7 till maintains compatibility with x32 that cannot use extended memory address space above the 4gb limit
Chack the two items of page two on this link especially the one about unchecking max memory use http://www.sevenforums.com/hardware-devices/5816-windows-7-64-bit-ram-issue-2.html

Not ruby programmer so leave that up to you, but for 3d vectors and using the general approach the angle is the shortest great circle path between the two. The angle is atan2( Norm( cross(a,b)), dot( a,b)). This means the angle is 0 to pi()
FYI http://www.euclideanspace.com/maths/algebra/vectors/angleBetween/index.htm and http://www.mathworks.com/matlabcentral/newsreader/view_thread/151925

If I make a circle, extrude that to cylinder, explode the bottom circle, select those edges via the edit menu , use Fredos Bezier plugin which has an polyline option then those edges are reported as curve entity( aka polyline per users guide?) in SU. I can weld those and if styles end points is selected you will note there is no intersection showing.Those can then be selected without putting the vertical lines in a differnt context.
Note if the vertical lines are keep as hidden then they are deleted during the above steps or some faces are lost ( when not hidden) which you need to remake.
Did I do something wrong or??
Using Su 7.1+ and have not checked for latest release of Freods plugin

Once one determines the angle value the inverse trig functions calculate the principal value. Therefore to determine the actual angle logic must be use either in the method or in the program its self to determine the correct quadrant. Give one can get the cos from the dot product, sine from the cross and even tangent from tan( theta) = |V1xV2|/ (V1 dot V2) then it is possible to establish the quadrant??

See if any ideas here help. Tried to find arctan2 for three 3d. More search probably required
http://www.euclideanspace.com/maths/algebra/vectors/angleBetween/index.htm

@jgb said:

@mac1 said:

jgb
have you tried this http://www.smustard.com/script/StrayLines

I think I looked at this some time ago, and found it not quite suitable in several, but common to me, occurrences. It works well with lines, but I had too many "no finds" when 2 or more vertexes of existing faces had a "no-face" micro gap between them. Maybe I used it wrong, dunno; too long ago.

I will try it again; thanks.

Applying multiple times helps some times and of course it does one seg at a time ,so you can make a stray line shorter and shorter
There are programs with filters that allow you to close close vertices based on a tolerance you input. Meshlab comes to mind but would not recommend you use. Check out ThmoThoms clean up since it may have a similar function. You may need to contact him to find out how to modiy code if it uses a fixed tol

jgb
have you tried this http://www.smustard.com/script/StrayLines

Petey
Form the look of your pic it appears there maybe surface and edge issues. Can you post your model to the 3d warehouse