NavList:

---

From: John Karl <jhkarl@att.net>
To: NavList <navlist@fer3.com>
Sent: Thu, November 5, 2009 6:51:52 PM
Subject: [NavList 10469] Re: No Two-Body Fix Problems


Peter,  OK, good.

I should also mention that there's no need to use the other solutions
to the inverse cosines in Eq. 7.5a&b (i.e., like 180-A) because those
solutions just switch (A+B) and (A-B) when substituted into cos(A-B)
in Eq. 7.5d (maybe this isn't obvious).  But that doesn't matter since
we're computing both solutions anyhow.

===========================

PH:
This detail was in fact quite clear.



===========================

And I agree that the equations should be perfectly well behaved in all
practical cases.  They only blow up at 90d declinations, 90d
altitudes, or zero D12 -- cases that will never occur in practice.
Also the equations will give the correct result for osculating LOPs.
In that case B=0, and the two solution for (A+B) and (A-B) will give
the same answer.

So these equations should be straight forward to implement in computer
code.

===========================

PH:
After a few iterations I arrived at an Excel spreadsheet showing that your equations are much more robust than you think.
D12=0 defines an ill-posed problem, so that is not a fair test.  However, even for declinations and altitudes of 90 degrees, osculating
LOPs (i.e. B=0 or 180), and below horizon sights (Ho<0), your procedure has passed all tests that I put it through so far.


Peter Hakel