Sun positions in solar studies are wrong

[clog boy]

How many of you make solar studies and sun/shadow animations as a contract piece? Revit has a nice function to calculate shadows based on time, date and location. Recently my Revit trainer told me that the positions are wrong and can't be calculated the right way.

For example, June 21st, the longest day of the year, there should be almost as much light at 6 in the evening as during, say, 2 in the afternoon. According to our sales department this is not the case; the rendering shows a nice sunset and the shadows are too long.

Did anyone else experience bad solar studies? Did anyone ever bother checking how accurate their solar studies are? You can understand this is a big issue that should be dealt with accordingly (i/e don't make decisions based on solar studies during the design process, demand from Autodesk it's fixed ASAP etc).

[trombe]

Hi Clog Boy.

I can say that I ( and certainly I expect all Revit users in New Zealand whether they be Architects or Draughtspeople or other Designers), use Revit sun shade / solar studies on contract, that is to say as part of paid work. It is a normal part of my work for what we call Resource Consent applications (land use planning permissions in general terms).

These are to show the "potential" affects both adverse and positive to surrounding land and buildings.

I have checked Revit 2008 edition and found it to be precise and correct. I have not checked Revit 2009.
I checked the shadows cast upon site plans and plan views, directly against 2 x A1 sheets (for each half of the year) which are full sheet sizes of the annual solar path for Wellington and detailed time for instance, down to minute increments etc.

These sheets were generated by a structural engineer on contract to the Wellington City Council in 1994 and handed to me by a council officer during resource consent deliberations in 1997.

I carefully compared angles and times over a set of about 20 diagrams and found every single one to be spot on.
It was a laborious task I can tell you but it did put my mind at rest because none of us want to be in a position where our claims as to accuracy could be at question. ( I printed stuff out and using protractors, parallel rulers and dividers as well as tracings sometimes, I ran guides and manual checks (also with hand calculator sometimes, other checking was also done within Revit using standard dimensions tools and overlays etc.))


That said, I do not know, if factory have changed anything at all when Accurender was removed and mental ray bound in in terms of the solar study tools / workings.
I assume not, but I have not checked once again.
Perhaps Steve Stafford would jump in here and confirm either way ?


Clearly, you need to ensure that you have your project and true north settings set up correctly to get the expected result, and this follows through to any other shadow casting procedures.
If you are sure that you have done all that stuff well and you feel it is still not right, then try a couple of small simple tests to check your results are consistently wrong and get peer review if possible to ensure you are not repeating a mistake and not realizing it.

Still not happy ? You should send a support request through to the factory.

regards
trombe

[clog boy]

Thanks.

I should review my initial post a bit (but am not going to edit it as it's thought provoking).
We have checked the project location, date and time settings and the True north, and all we did was redefine our origin. This time we got better results but still can't make solid claims of it being right.

We are not able to say whether it's 'a bit off', 'completely off' or '100% right'. As it stand now each of the three statements is as likely as the other until we are able to simulate an existing situation.

All we can do is make a mass model of our office, make a rendering the moment it's a sunny day with the date- and time settings of that moment, and go outside to see whether it's reliable or not. Then reverse-engineer some settings to implement in our templates.

Though it's likely that redefining the origin resolved the issue.

What settings did you use to calculate your shadows?

[trombe]

Ho CB,
settings?
To clarify, my project was not linked or part of a workset, but a stand alone project in a single file, so there were no possible issues to do with locations.
Under Advanced Model Graphics_Sun and Shadow Settings_Name = Single Day_Place = Lat = -41.303 Long 174.7372.
these were the primary settings for place.
What I do usually is set up a series of specific days per year and usually for 4 winter and 4 summer months, and check times when I expect the critical issues to become manifest (for example, but often these selected times and days vary widely)

The study is often for but for a single day to run a complete day study and then instead of exporting it as an avi file, I tell Revit to export as jpeg file. So then I get a set of nominated still shots per time period.
(One project where there was a school as a neighbour, required about 64 images to cover off all of the possibilities)

I do this for existing and then for proposed, export them out and then import them again to lay up on sheets side by side for easier direct visual comparison.
So having these on plan, is a simple and accurate way to be able to compare what you can check off with a separate source / data set.
Is this in the ball park for information ?

regards
trombe

[iankids]

Hi All,

The accuracy of the shadows is a huge concern as often during the town planning process matters end up before tribunals where these sorts of things are poured over in painstaking detail.

Thus in the first instance, the paramount importance to me is not so much how the shadows relate to real life, but rather how the shadows as shown on my drawings relate to the well established guidelines for shadows within Melbourne.

For well over a decade or more, all designers, town planners and others have used the same template for drawing and or assessing shadow impacts.

This template states that for a 1.0 mt high post at:

9:00am the shadow is 1.6mts
10:00am the shadow is 1.15mts
11:00am the shadow is 0.87mts
12:00 noon the shadow is 0.78mts
1:00pm the shadow is 0.84mts
3:00pm the shadow is1.3mts

I set up a quick test using a 1.0mt high wall and the results were:

9:00am - 1.63mt
11:00am - 0.9mts
12 noon - 0.79mts
1:00pm - 0.84mts
3:00pm - 1.33mts

Whilst there are some slight variances from the accepted table, the differences are at least on the high side, and by and large I would not view them as a material difference.

Do these "accurately" reflect real life? Not so sure as the subtle vagaries of topography etc comes into play.

But from a design / application process point of view, I am quite happy to stand up in court and declare that they are a true and reasonable representation of the likely shadow impact.

Cheers,

Ian

[clog boy]

Thanks, I'm now quite confident that a solar study, given the right settings such as time, place and orientation, are a reliable representation of real life and not something bodged together for eyecandy's sake.

That's, ofcourse, assuming that if it works 'down under' it should also work halfway across the globe here in The Netherlands. I'm going to look for some reference material and see if it's right.

By the way, how did you meassure the shadows in Revit?

[iankids]

By the way, how did you meassure the shadows in Revit?

The dirty simple way - draw a detail line from the wall along the edge of the shadow & measure it!?!

If there is a better way out there, I would love to know about it.

Cheers,

Ian

[gordolake]

I believe there is a problem, at least for the southern hemisphere.
My project is setup correctly with north aligned to true north by survey data. The shadow in a plan oriented 3D view is calculating the shadows incorrectly. To get them right I have to subtract the rotated true north value from 90 degrees and apply that for it to be correct i.e. if the project is 34 degrees west for true north I have to set it to 56 degrees west.

This must be addressed in the next build as shadow diagrams form part of the approval process and as designers we would look really stupid submitting these results to the authorities.

Hope this helps others with some of the hair pulling.

Steve

[charliep]

So how do you determine true north?

[patricks]

So how do you determine true north?

Most CAD surveys are drawn to true north. I recently had one that was off quite a bit, as determined by the site viewed on Google Earth. I called up the engineer and he got me a new file that was rotated correctly.

As for the shadow studies, you could always look up some solar angle information for your area (there are probably calculators online that can tell you altitude and azimuth at a given lat/long), do a little math and see if your calculations match Revit's.

[narlee]

I have found that the early evening setting are way too dark, so that I have to "cheat" the time of day.

[trombe]

So how do you determine true north?

charliep,


There are a number of reliable ways to get this information.
You could - ask your local council officers in the planing department.
OR contact your local meteorological office or university science department - astrophysics or physics....this information could also be on the web for your locaiton.

For me, I got my figures from our ESR, which had been the national DSIR - Department of Scientific and Industrial Research ....but any weather service has this accurate data for every location down to points of a second. The local variation is usually small but in constant flux with the magnetic field of the earth...no court is gonna hang ya butt on the semantics !

cheers
trombe

[ArchUK]

I am based in the UK and we have been asked to provide a BS (British Standard) shadow study. We normally can provide a shadow study using the revit software but i am aware and have seen and commissioned BS Study's too.

I personally have advised the client i will not be providing a BS Planning Sunlight, Daylight and Overshadowing Report as I can not access any technical literature from Autodesk that reveals how it commissions the relevant information beyond... select the location and run it.

For example the BRE (Building Research Establishment)... advise the following:

For sunlight: The sun light protractor method and sunlight availability indicator for 51.5o N as set out in Appendix A of SLP.

For daylight VSC: The principles set out in section 2 of SLP together with the concept of Vertical Sky Component (VSC%) as set out in both Appendix A of SLP – and in BS 8206-2:2008:code of practice for daylighting.

For shadow paths: The proposals are digitally modelled in Integrated Environmental Solutions’ (IES) Model IT software and then analysed in IES suncast, version v6.4. Shadows are predicted at hourly intervals on the equinox date, 21st March, in accordance with the BRE criteria.

So, you see why the google your location and your good to go approach leaves me with slight hesitancy on standing behind revits accuracy.

That said... i did the 1m post test as the contributor above advised and the results practically the same (or close enough to be of no material impact).

I would appreciate a little more from autodesk on it though as its a hugely beneficial tool for me!