qheat did not converge pah grain with error code 14 in PN model (trac #82)

[cloudy-bot]

reported by: peter

With the attached input script the stochastic heating code produces lots of messages like this:

PROBLEM  qheat did not converge grain  pah-ab0803 in zone 588, error code = 14

This is because the model is spherical and goes very deep, which causes the radiation field to become very diluted making heating events of the grains very rare. This then allows a sizable fraction of the grains to cool below GRAIN_TMIN, which in turn causes the error messages.

The solution is to relax the error checks when qtmin is very close to GRAIN_TMIN since we don't really care about these extremely cool grains. They contribute very little to the flux and only at very long radio wavelengths (longwards of 1 meter) where the radiation will likely be drowned by free-free emission from the electrons further in.

In order to reproduce the problem a modified version of dense_fabden() is needed, which is also attached.

Problem reported by Patrizia Manzitto

Migrated from https://www.nublado.org/ticket/82

{
    "status": "closed",
    "changetime": "2009-02-05T15:22:28Z",
    "_ts": "1233847348000000",
    "description": "With the attached input script the stochastic heating code produces lots of messages like this:\n\n{{{\nPROBLEM  qheat did not converge grain  pah-ab0803 in zone 588, error code = 14\n}}}\n\nThis is because the model is spherical and goes very deep, which causes the radiation field to become very diluted making heating events of the grains very rare. This then allows a sizable fraction of the grains to cool below GRAIN_TMIN, which in turn causes the error messages.\n\nThe solution is to relax the error checks when qtmin is very close to GRAIN_TMIN since we don't really care about these extremely cool grains. They contribute very little to the flux and only at very long radio wavelengths (longwards of 1 meter) where the radiation will likely be drowned by free-free emission from the electrons further in.\n\nIn order to reproduce the problem a modified version of dense_fabden() is needed, which is also attached.\n\nProblem reported by Patrizia Manzitto",
    "reporter": "peter",
    "cc": "",
    "resolution": "fixed",
    "time": "2009-02-05T14:35:58Z",
    "component": "grains",
    "summary": "qheat did not converge pah grain with error code 14 in PN model",
    "priority": "minor",
    "keywords": "",
    "version": "",
    "milestone": "",
    "owner": "peter",
    "type": "defect - convergence"
}

[cloudy-bot]

• @peter-van-hoof-noaccount changed attachment from "" to "patrizia2.in"
• @peter-van-hoof-noaccount commented:

input script

[cloudy-bot]

• @peter-van-hoof-noaccount changed attachment from "" to "dense_fabden.cpp"
• @peter-van-hoof-noaccount commented:

modified dense_fabden()

[cloudy-bot]

@peter-van-hoof-noaccount changed description from:

With the attached input script the stochastic heating code produces lots of messages like this:

PROBLEM  qheat did not converge grain  pah-ab0803 in zone 588, error code = 14

This is because the model is spherical and goes very deep, which causes the radiation field to become very diluted making heating events of the grains very rare. This then allows a sizable fraction of the grains to cool below GRAIN_TMIN, which in turn causes the error messages.

The solution is to relax the error checks when qtmin is very close to GRAIN_TMIN since we don't really care about these extremely cool grains. They contribute very little to the flux and only at very long radio wavelengths (longwards of 1 meter) where the radiation will likely be drowned by free-free emission from the electrons further in.

In order to reproduce the problem a modified version of dense_fabden() is needed, which is also attached.

to:

With the attached input script the stochastic heating code produces lots of messages like this:

PROBLEM  qheat did not converge grain  pah-ab0803 in zone 588, error code = 14

This is because the model is spherical and goes very deep, which causes the radiation field to become very diluted making heating events of the grains very rare. This then allows a sizable fraction of the grains to cool below GRAIN_TMIN, which in turn causes the error messages.

The solution is to relax the error checks when qtmin is very close to GRAIN_TMIN since we don't really care about these extremely cool grains. They contribute very little to the flux and only at very long radio wavelengths (longwards of 1 meter) where the radiation will likely be drowned by free-free emission from the electrons further in.

In order to reproduce the problem a modified version of dense_fabden() is needed, which is also attached.

Problem reported by Patrizia Manzitto

[cloudy-bot]

• @peter-van-hoof-noaccount commented:

Fixed in r2695 on the trunk.

Fixed in r2696 on c08_branch.

• @peter-van-hoof-noaccount changed resolution from "" to "fixed"
• @peter-van-hoof-noaccount changed status from "new" to "closed"