This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 1 Atms 502 Numerical Fluid Dynamics Thu., Oct. 12, 2006 10/18/06 Atms 502  Fal 2006  Jewett 2 Test case Cone • Upstream method • Refinement: 3x True Feedback No feedback 10/18/06 Atms 502  Fal 2006  Jewett 3 True Feedback No feedback Test case Square / top hat • LaxWendroff • Refinement: 3x Looks a bit ugly Error 50% less than upstream Note nest only on right half of square 10/18/06 Atms 502  Fal 2006  Jewett 4 Update frequency Below: error scores vs update frequency Frequent updates  lower error Less frequent updates  increasing error Hmm… 2 10/18/06 Atms 502  Fal 2006  Jewett 5 Oscillations • How long for features to move 1 ∆ x ??? s=vt ; time = distance/speed; t=s/v Total time to go distance ∆ x = ∆ x/c # time steps to go ∆ x = 1/ ν (why?) • So, every 5 time steps … The cone peak moves one grid distance. What does this say about our “true” solution? 10/18/06 Atms 502  Fal 2006  Jewett 6 Oscillations • Every 5 time steps the (exact) solution moves one grid length. • At right: 2 periods in 10 steps (1) What are these oscillations? • Projecting an exact solution on a finite grid results in errors  in the “true” solution !! n N+2 N+4 N+6 N+8 N+10 10/18/06 Atms 502  Fal 2006  Jewett 7 Question 3: update frequency Below: error scores vs update frequency Program 2b  cone  LaxWendroff Frequent updates  lower error Less frequent updates  increasing error total error after 200 steps  NX=70 case 10/18/06 Atms 502  Fal 2006  Jewett 8 Question 3: update frequency • In 60 steps the middle of the cone is on the right nest domain edge. • In 30 steps , when the error starts rapidly rising, the right edge of the cone (and of trunc. error) hits the nest edge. • In 20 steps , the nest border is still ahead of the cone’s leading edge. 60 steps 30 steps 20 2 10/18/06 Atms 502  Fal 2006  Jewett 5 Oscillations • How long for features to move 1 ∆ x ???...
View
Full
Document
This note was uploaded on 10/06/2009 for the course ATMOSPHERI 502 taught by Professor Jewett during the Spring '09 term at University of Illinois at Urbana–Champaign.
 Spring '09
 JEWETT

Click to edit the document details