Thunderhead Engineering Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Forum moved to https://forum.thunderheadeng.com

Author Topic: Calculate comfort distance <-> p/m²?  (Read 3172 times)

Robin

  • New Member
  • *
  • Posts: 6
    • View Profile
Calculate comfort distance <-> p/m²?
« on: June 03, 2014, 12:39:43 am »

Hi,

I´m working in steering mode and I want to determine the comfort distance to be able to reach a maximum of 3 persons / m² when they are queuing.

- I´m using the standard size of occupants (45,58 cm).

Is there a formula?
I have been reading the technical manual without finding any clear result.

Thanks in advance

Best Regards
Robin Averstedt
Robin.averstedt@ramboll.se
Logged

Richard O'Konski

  • Thunderhead
  • *****
  • Posts: 28
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #1 on: June 03, 2014, 11:02:27 am »

Robin,

The way you would affect the density in Pathfinder would be to set the comfort distance between occupants (Model->Edit Profiles...->Advanced). We do not currently have a formula that describes the relationship between comfort distance and density, however, which means you will need to do some trial and error, but you could use this to find almost any density.

After performing the following steps, I came up with a comfort distance of .08 m to achieve a density of 3 persons/m2:
  • Open the attached model in Pathfinder 2014.
  • Calculate an upper bound on the comfort distance using the following formula:
    C=1.422/D.5 - W
    where C is the comfort distance, D is your required density, and W is the shoulder width of an occupant. This formula assumes that occupants arrange themselves in a tight pattern (sphere packing). Occupants in Pathfinder don't actually pack this tightly, but this will give you an upper bound on the comfort distance.
  • Set this as the comfort distance for the default occupant profile.
  • Run the simulation.
  • View Room Usage Results (Results->View Room Usage...)
  • Uncheck the box next to "Remaining (Total)" and check the box next to "Floor 0.0 m->Room00".
  • Take the peek number of occupants and divide it by 100 (the area of that room).
  • If this doesn't match the required density, adjust the comfort distance and repeat steps 4-8.

Logged
Richard O'Konski
Thunderhead Engineering

Robin

  • New Member
  • *
  • Posts: 6
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #2 on: June 04, 2014, 01:07:20 am »

Thanks Richard.

So, the formula is for an upper limit, correct?

c=(1.422/(D^0.5))-W
c=(1.422/(3^0.5))-0.4558=0.3652?

Just checking, have I interpreted the formula right?!


Can you also answer a supplymentary question?

- Does the comfort zones for different occupants overlap each other or are they possibly programmed like it´s shown in figure 9a-b, at page 28 in the technical manual?


Cheers

//Robin
Logged

Richard O'Konski

  • Thunderhead
  • *****
  • Posts: 28
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #3 on: June 04, 2014, 08:17:55 am »

Robin,

You have interpreted the formula correctly. You will most likely need to use a lower value for comfort distance than the calculated value to achieve your desired density. For instance, after trial and error I came up with the value of .08 for comfort distance to achieve 3 pers/m2.

I'm not quite sure I understand your second question.  The comfort distance is the distance an occupant tries to maintain between his/her body and another occupant's body.  Figure 9 in the technical manual is illustrating two different configurations in which occupant's might have conflicting paths and is not necessarily related to comfort distance.

If you want to learn more about how the comfort distance affects decision making, see the section, "Avoid Occupants", of the technical manual.
Logged
Richard O'Konski
Thunderhead Engineering

Robin

  • New Member
  • *
  • Posts: 6
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #4 on: June 05, 2014, 12:47:50 am »

Please see the attached pdf as an explanation to my question.

I´m thinking that it’s easier for the programmers to handle the comfort distance as in figure 1.
But, perhaps it´s handled like in figure 2, but then it must be hard to handle it when a occupant is standing beside a wall? You are not defining two different comfort distances?

The illustrated picture I referred to in the technical manual, it seems like the comfort distance is as in figure 1.

Do you know, out of those figures, how it´s defined?

Thanks for all replies!

Best regards
Robin Averstedt
Logged

Richard O'Konski

  • Thunderhead
  • *****
  • Posts: 28
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #5 on: June 05, 2014, 08:15:59 am »

The comfort distance is implemented as shown in Figure 2 of your image.  Each occupant is programmed to be autonomous, which means that he/she has no knowledge of the other occupant's comfort distance as would be needed in Figure 1, since each occupant can have a different comfort distance. 

In addition, my earlier description of the comfort distance may be a bit misleading.  Occupants don't actually try to maintain this distance away from walls/occupants. Instead, as an occupant is moving, they calculate intersections with other occupants/walls along their path. If they sense that they will hit the other occupant/wall within this distance, they stop.  This means that one occupant can be travelling directly behind another occupant at the same speed and within a distance smaller than the comfort distance because they will not detect that they will hit one another. The same is true for walls. The occupant can travel right alongside a wall because they won't detect an intersection with that wall. This is a feature we hope to improve in the future.
Logged
Richard O'Konski
Thunderhead Engineering

Robin

  • New Member
  • *
  • Posts: 6
    • View Profile
Re: Calculate comfort distance <-> p/m²?
« Reply #6 on: June 05, 2014, 08:27:53 am »

Thank you very much for the clarification!
Logged