PDA

View Full Version : General Revit Workflow



rjcrowther
2006-12-20, 08:57 AM
Has anyone come across a generalised procedure for building a model? (Realistically it would be Residential - Commercial would perhaps have too many variables) I am thinking there is generally a set method for building the real thing so should one also adopt the same approach when modeling?

For some background, I have just done a Residential Energy Rating Course and the software manufacturer has laid down a set procedure to model the spaces, massing, overhangs, etc so nothing is missed. This prompted me to this question.

Thanks,
Rob

Justin Marchiel
2006-12-20, 03:19 PM
i am not sure how you work, but for us, we generally start with a plan to show the client. from there we evolve it to a point where we are confortable and then start looking at the elevations and sections.

I don't know if i understand the questions correctly, but i find the workflow is more of model it when needed approach. Time is money, and i dont see a point adding roofs, floors, structure to a model when it is not called for. some of my users get boged down trying to sort out details in the schematic phase. i tell them that it is really not important because the design will change and the time you spend modeling some information is lost.

For me we only spent time on creating containt when required. Then at each stage of the process more and more information is added into the model. no use have all that extra stuff in there, especially if the job gets cancelled. i fell better about a job when it gets cancelled (as good as you can fell when a job is cancelled) because we only have the information that was required to show the client. so we get paided fully for what we have shown and did waste time and money on information that we could not bill.

Justin

david.metcalf
2006-12-20, 05:04 PM
I think he meant the process of taking the model to some point where coordination with engineers and outside consultants on the model is needed and how is that data from the model exchanged.

It really depends. Who are the consultants, what are they using for software, what kind of format is required from each consultant, etc. The process work flow can be fairly extensive. For example your LEED consultant may only need a 2d file in 2000/2005/2007 DWG format or may need a 3d mass model of the spaces to extract information from. This is where you need to do a assessment of your process and the processes of the consultants to document the work flow.

Is this is where you want to go?

rjcrowther
2006-12-20, 11:09 PM
Thanks for the replies.

I wasn't thinking so much in terms of outside consultants although this is factor.

I was thinking along the lines of somehow formalising what Justin is doing which is what I do, and that is jumping around the model at various stages and adding bits and pieces as they arise.

Certainly I don't want to model too much in the beginning either. There comes at point when the decision is made that the the project will become a reality and it is from here that I am thinking a set procedure may be of benefit.

After approval on the design is given, my method is to model in a way the will allow the resulting 2D drawings to meet with my drawing checklists.
A down side is in this example:
I generally do not add cornices to internal walls but just draw them. In a recent extension a fair time was spent with filled regions and groups adding cornices. Should have modeled them (wall sweeps) - it would have been easier and it removes the prospect of 2D drafting errors.

From this I was wondering if a generalised procedure would help, something like construction scheduling. After writing this, and reading the replies, I think it is probably an individual effort of setting your own guidelines as to the extent of modeling required to satisfy the project (and yourself) and make a checklist to make sure everything that should be modeled has been. Then just draw the rest.

Rob

ron.sanpedro
2006-12-20, 11:59 PM
Has anyone come across a generalised procedure for building a model? (Realistically it would be Residential - Commercial would perhaps have too many variables) I am thinking there is generally a set method for building the real thing so should one also adopt the same approach when modeling?

For some background, I have just done a Residential Energy Rating Course and the software manufacturer has laid down a set procedure to model the spaces, massing, overhangs, etc so nothing is missed. This prompted me to this question.

Thanks,
Rob

For what it is worth, I am finding that there are two very different work flows that I am running into. The first is very similar to the energy modeling example. I have a known design, and I am modeling it. In this case, I almost always start at plan and do 90% of my work there. I might reference the elevation, be it someone else's DWG shell drawings, or a designers sketch, or whatever. But I don't set window height and head height in elevation, I do it in plan. I will reference a building section to understand construction, but I will model most all of it in plan again. Even roof slope, I look at the reference elevation, see that I have 4:12, and I plug that in while doing the roof, usually in the site plan because I don't even bother to make a roof plan view. Often I am modeling an existing shell for a TI and there will never be a Roof plan on a sheet, or a building section for that matter. The process is very plan oriented in Revit, and very minimalist as far as views created.

The other work flow is when Revit is the design tool. There I might just drop a bunch of windows in plan, set a width that makes sense, then actually set heights, and even change head heights, in the elevation where I can make a judgment call. Then I will shoot an interior elevation and make sure it works there too. When I model the roof I might actually shoot a perspective from the street, and play around with roof slopes from there. I might cut a ton of sections that last just long enough to check a stair head hight, or make an evaluation of a split level condition, then delete it again to keep the clutter down. When Revit is a design tool the shadows get turned on, and materials added, and schedules that will never make it on a sheet but that tell me things like how much demo wall do I have in this scheme. All stuff that I would never bother with when I am simply modeling a thing that is already defined.

As for your mention of what to model and what not to, I am leaning towards modeling it, but in only as much detail as needed. And also, will I need it in another view, if I decide to use that view. So, in the case of a picture rail or trim or the like, I would model it. A simplified profile to be sure, but still, I want it there in that section I didn't think I would need but found that I did. And the interior elevation I didn't plan on doing but the client was struggling with how the paint colors would really relate, etc. However, I don't model the flashing for an elevation, because I won't need it in 3D either. And if I do, I "model" it as a line, not an actual item, because that woks in elevation and 3D, and in section I can ignore it until the detail, where a line again works better.

And I am finding that TIs really change what gets modeled. We had a single floor TI of a three story building, and we figured a good twenty hours to "model the existing conditions". Of course this was based on doing the whole building, with all wall types defined correctly, etc. Instead I have a wall type for each overall thickness of wall, I modeled nothing but the third floor (you look at an elevation and the stairs hang out the bottom of the building), and I kept everything to a minimum. Six hours later we have existing conditions to do a TI with.

So even when working with a known thing, the process can vary with project type and scope. And some of that should be consistent no matter who is actually doing the work, and indeed is being documented as TI Existing Conditions Guidelines for Revit.

Gordon

rjcrowther
2006-12-21, 01:06 AM
Thats quite a detailed reply, thanks for you efforts Gordon. it is great to read of your methods and thoughts.

I am also interested to see you are leaning towards modelling something than not (I take note about level of detail). It is that unexpected detail or section that can become a pain in the neck.

Rob