[mesa-users] approx13
Ken Shen
kenshen at astro.berkeley.edu
Wed May 14 17:57:36 EDT 2014
Hi Bill, performance isn't an issue for my actual calculations, so I've
been using soft-wired networks with lots of isotopes. But I'm trying to
compare to someone else's calculation, which used approx13...
On Wed, May 14, 2014 at 2:54 PM, Bill Paxton <paxton at kitp.ucsb.edu> wrote:
> Hi Ken,
>
> Hardwired nets are a challenge! It is much easier to screw them up than
> to get them right. But if you really need to squeeze out the last bit of
> performance, then they are required. So, as background, is approx20.netjust too big for you to use? Really? If so, then it will be possible
> (with some work on my part) to create a smaller net like frank's approx13.
> let me know.
>
> -b
>
>
>
>
>
> On May 14, 2014, at 2:09 PM, Ken Shen wrote:
>
> Hi all. I've been trying to construct a nuclear network in MESA that
> mimics Frank Timmes's approx13, which is an alpha-network that contains
> 4He, 12C, 16O, 20Ne, etc. up to 56Ni. It contains the 12x2 (a,g) and (g,a)
> forward and reverse reactions linking each adjacent isotope (e.g.,
> 12C(a,g)16O); 3 heavy-ion forward reactions (12C+12C, 12C+16O, 16O+16O);
> and 8x4 (a,p)(p,g) forward and reverse p-catalyzed alpha-captures starting
> at 24Mg(a,p)(p,g)28Si. These hardwired reactions assume the intermediate
> isotopes (e.g., 27Al) are in equilibrium with the reactions that involve
> them.
>
> MESA (version 5596) gives me an error when I don't include protons in the
> network (why is that?), so I've added them in, but no reactions link
> directly to them, so their mass fraction should stay zero (and does). I've
> attached the naive .net file I came up with, based off approx_base.net,
> at the bottom of the email.
>
> The problem is that this network doesn't give the correct burning results.
> At T = 3e9 K and rho = 1e7 g/cm^3, a pure 12C composition results in
> epsilon = 1.8e19 erg/(g*s), which is about 4 orders of magnitude too low.
> I eventually stumbled across a fix, which is to include neutrons in the
> isotope list. As with protons, neutrons don't show up explicitly in the
> network and their mass fraction stays constant at zero, but the burning
> goes up to the correct value.
>
> Furthermore, with neutrons and protons included in the isotopes list, a
> pure 16O composition at the same T and rho yields epsilon = 6.3e17. Adding
> 1H to the list (again, not part of the actual network) yields epsilon =
> 7.3e17.
>
> Any enlightenment out there?
>
>
> ! Naive approx13.net file
> add_isos(
> prot
> he4
> c12
> o16
> ne20
> mg24
> si28
> s32
> ar36
> ca40
> ti44
> cr48
> fe52
> ni56
> )
>
> add_reactions(
>
> ! triple alpha
> r_he4_he4_he4_to_c12
> r_c12_to_he4_he4_he4
>
> ! C/O burning
> r1212 ! c12 + c12 -> ne20 + he4, includes na23+p
> channel
> r1216_to_mg24 ! c12 + o16 -> mg24 + he4
> r1216_to_si28 ! c12 + o16 -> si28
> r1616a ! o16(o16a)si28
> r1616ppa ! o16(o16p)p31(pa)si28
> r1616ppg ! o16(o16p)p31(pg)s32
> r1616g ! o16(o16g)s32
> r1616p_aux ! o16(o16p)p31
>
> ! alpha links with both (a,g) and (a,p)(p,g)
>
> ! o16
> ! from c12 by (a,g)
> r_c12_ag_o16
> r_o16_ga_c12
>
> ! ne20
> ! from o16 by (a,g)
> r_o16_ag_ne20
> r_ne20_ga_o16
>
> ! mg24
> ! from ne20 by (a,g)
> r_ne20_ag_mg24
> r_mg24_ga_ne20
>
> ! si28
> ! from mg24 by (a,g)
> r_mg24_ag_si28
> r_si28_ga_mg24
>
> ! from mg24 by (a,p)(p,g) via al27
> rmg24ap_to_si28 ! mg24(ap)al27(pg)si28
> rsi28gp_to_mg24 ! si28(gp)al27(pa)mg24
> rmg24ap_aux ! mg24(ap)al27
> ral27pg_aux ! al27(pg)si28
> rsi28gp_aux ! si28(gp)al27
> ral27pa_aux ! al27(pa)mg24
>
> ! s32
> ! from si28 by (a,g)
> r_si28_ag_s32
> r_s32_ga_si28
>
> ! from si28 by (a,p)(p,g) via p31
> rsi28ap_to_s32 ! si28(ap)p31(pg)s32
> rs32gp_to_si28 ! s32(gp)p31(pa)si28
> rsi28ap_aux ! si28(ap)p31
> rp31pg_aux ! p31(pg)s32
> rs32gp_aux ! s32(gp)p31
> rp31pa_aux ! p31(pa)si28
>
> ! ar36
> ! from s32 by (a,g)
> r_s32_ag_ar36
> r_ar36_ga_s32
>
> ! from s32 by (a,p)(p,g) via cl35
> rs32ap_to_ar36 ! s32(ap)cl35(pg)ar36
> rar36gp_to_s32 ! ar36(gp)cl35(pa)s32
> rs32ap_aux ! s32(ap)cl35
> rcl35pg_aux ! cl35(pg)ar36
> rar36gp_aux ! ar36(gp)cl35
> rcl35pa_aux ! cl35(pa)s32
>
> ! ca40
> ! from ar36 by (a,g)
> r_ar36_ag_ca40
> r_ca40_ga_ar36
>
> ! from ar36 by (a,p)(p,g) via k39
> rar36ap_to_ca40 ! ar36(ap)k39(pg)ca40
> rca40gp_to_ar36 ! ca40(gp)k39(pa)ar36
> rar36ap_aux ! ar36(ap)k39
> rk39pg_aux ! k39(pg)ca40
> rca40gp_aux ! ca40(gp)k39
> rk39pa_aux ! k39(pa)ar36
>
> ! ti44
> ! from ca40 by (a,g)
> r_ca40_ag_ti44
> r_ti44_ga_ca40
>
> ! from ca40 by (a,p)(p,g) via sc43
> rca40ap_to_ti44 ! ca40(ap)sc43(pg)ti44
> rti44gp_to_ca40 ! ti44(gp)sc43(pa)ca40
> rca40ap_aux ! ca40(ap)sc43
> rsc43pg_aux ! sc43(pg)ti44
> rti44gp_aux ! ti44(gp)sc43
> rsc43pa_aux ! sc43(pa)ca40
>
> ! cr48
> ! from ti44 by (a,g)
> r_ti44_ag_cr48
> r_cr48_ga_ti44
>
> ! from ti44 by (a,p)(p,g) via v47
> rti44ap_to_cr48 ! ti44(ap)v47(pg)cr48
> rcr48gp_to_ti44 ! cr48(gp)v47(pa)ti44
> rti44ap_aux ! ti44(ap)v47
> rv47pg_aux ! v47(pg)cr48
> rcr48gp_aux ! cr48(gp)v47
> rv47pa_aux ! v47(pa)ti44
>
> ! fe52
> ! from cr48 by (a,g)
> r_cr48_ag_fe52
> r_fe52_ga_cr48
>
> ! from cr48 by (a,p)(p,g) via mn51
> rcr48ap_to_fe52 ! cr48(ap)mn51(pg)fe52
> rfe52gp_to_cr48 ! fe52(gp)mn51(pa)cr48
> rcr48ap_aux ! cr48(ap)mn51
> rmn51pg_aux ! mn51(pg)fe52
> rfe52gp_aux ! fe52(gp)mn51
> rmn51pa_aux ! mn51(pa)cr48
>
> ! ni56
> ! from fe52 by (a,g)
> r_fe52_ag_ni56
> r_ni56_ga_fe52
>
> ! from fe52 by (a,p)(p,g) via co55
> rfe52ap_to_ni56 ! fe52(ap)co55(pg)ni56
> rni56gp_to_fe52 ! ni56(gp)co55(pa)fe52
> rfe52ap_aux ! fe52(ap)co55
> rco55pg_aux ! co55(pg)ni56
> rni56gp_aux ! ni56(gp)co55
> rco55pa_aux ! co55(pa)fe52
>
>
> )
>
> ------------------------------------------------------------------------------
> "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
> Instantly run your Selenium tests across 300+ browser/OS combos.
> Get unparalleled scalability from the best Selenium testing platform
> available
> Simple to use. Nothing to install. Get started now for free."
>
> http://p.sf.net/sfu/SauceLabs_______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20140514/50ac29c8/attachment.html>
More information about the Mesa-users
mailing list