The Student Room Group

Using corrected value for hardy cross method

i have several example of hardy cross problem involve 2 loops , for the junction that divide 2 loops , is it necessary to use the corrected value for the junction on second loop for the first trial ? or just use the initial assumption will do ?

http://nptel.ac.in/courses/112104118/lecture-36/hyperlink/examp_36_1.htm

using this example , the BD is the junction that divide 2 loops , should we use the corrected value 0.8 -(-0.46) =1.26 or stick to the original assumption (0.8) just like the author ?

in the second example here , the author use the corrected value (0.01731) for junction BE in the second loop instead of 0.0

http://i.imgur.com/RWZ5odq.png

Scroll to see replies

Original post by wilson dang
i have several example of hardy cross problem involve 2 loops , for the junction that divide 2 loops , is it necessary to use the corrected value for the junction on second loop for the first trial ? or just use the initial assumption will do ?

http://nptel.ac.in/courses/112104118/lecture-36/hyperlink/examp_36_1.htm

using this example , the BD is the junction that divide 2 loops , should we use the corrected value 0.8 -(-0.46) =1.26 or stick to the original assumption (0.8) just like the author ?

in the second example here , the author use the corrected value (0.01731) for junction BE in the second loop instead of 0.0

http://i.imgur.com/RWZ5odq.png


Disclaimer: Haven't actually used Hardy-Cross - my fluid dynamics course was much more GFD oriented than engineering oriented

From what I understand the initial guess values must satisfy Kirchoff's laws (so the guess values in the first example I believe are incorrect as they only satisfy the current law - though it should still converge). Since it is an iterative method, with both loops being iterated simultaneously, the uncorrected value should be used in both loops in the first iteration - it is likely that the solution will still converge either way but may take longer one way.

Will try both and get back to you :smile:
Original post by wilson dang
i have several example of hardy cross problem involve 2 loops , for the junction that divide 2 loops , is it necessary to use the corrected value for the junction on second loop for the first trial ? or just use the initial assumption will do ?

http://nptel.ac.in/courses/112104118/lecture-36/hyperlink/examp_36_1.htm

using this example , the BD is the junction that divide 2 loops , should we use the corrected value 0.8 -(-0.46) =1.26 or stick to the original assumption (0.8) just like the author ?

in the second example here , the author use the corrected value (0.01731) for junction BE in the second loop instead of 0.0

http://i.imgur.com/RWZ5odq.png


Update: doesn't really matter which you do, using the corrected value seems to have slightly faster convergence but the iterative step is slightly more complicated.

For the first example, using the uncorrected value gives a result on the fifth iteration but using the corrected one only takes four iterations.
Reply 3
Original post by natninja
Update: doesn't really matter which you do, using the corrected value seems to have slightly faster convergence but the iterative step is slightly more complicated.

For the first example, using the uncorrected value gives a result on the fifth iteration but using the corrected one only takes four iterations.


http://i.imgur.com/RWZ5odq.png

if i am using the corrected flow rate for junction BE at the second loop ? is it necessary to make changes to the other flow rate at other junction so that they are 'balance'
e.g. in the initial assumption , 100L/s flow into a node , and there should be 50, 50 flow out or 30 and 70 flow out .

if i found the corrected value is 25 at one of the junction , is it necessary to make changes to the other flow rate to the flow rat eat another junction ( so now become 25 , and 75 flow out ) ?
P/s : in the example , i found that the author didnt make changes to the other flow rate , is it correct ?
So I'm not really sure what you are trying to say but here's a breakdown in terms of the iteration:

Using uncorrected value:
Iterate all loops simultaneously with one set of values

Using corrected value:
Iterate first loop
Obtain corrected value for junction
iterate second loop using corrected value obtained from first loop without modifying any other flow rate
Obtain corrected values for all pipes

So no, you don't have to make changes to the other flow rate to balance them though you probably can and it would probably speed it up. I'll try it and get back to you
Reply 5
Original post by natninja
So I'm not really sure what you are trying to say but here's a breakdown in terms of the iteration:

Using uncorrected value:
Iterate all loops simultaneously with one set of values

Using corrected value:
Iterate first loop
Obtain corrected value for junction
iterate second loop using corrected value obtained from first loop without modifying any other flow rate
Obtain corrected values for all pipes

So no, you don't have to make changes to the other flow rate to balance them though you probably can and it would probably speed it up. I'll try it and get back to you


do let me know is it necessary to make changes to the other flow rate to balance them later ....
Original post by wilson dang
do let me know is it necessary to make changes to the other flow rate to balance them later ....


Can confirm, do not do this - it causes the iterative process to diverge so you would never reach a solution...
Reply 7
Original post by natninja
Can confirm, do not do this - it causes the iterative process to diverge so you would never reach a solution...

So, it is not necessary to make any changes to the flow rate atq the other junction?? Just use the initial assumption for the first trial on the second loop?

Sent from my C6502 using Tapatalk
Original post by wilson dang
So, it is not necessary to make any changes to the flow rate atq the other junction?? Just use the initial assumption for the first trial on the second loop?

Sent from my C6502 using Tapatalk


You can use the corrected value for the junction but do not attempt to rebalance the other flow rates during the iteration
Reply 9
Original post by natninja
You can use the corrected value for the junction but do not attempt to rebalance the other flow rates during the iteration

How about the second trial of first loop of BE? should I use the corrected value from BE in the first loop or second loop during the first trial??

Sent from my C6502 using Tapatalk
Original post by wilson dang
How about the second trial of first loop of BE? should I use the corrected value from BE in the first loop or second loop during the first trial??

Sent from my C6502 using Tapatalk


Only use corrected values for loops after the correction has been done.

It's simpler to use the uncorrected value, even if it could take an additional iteration or two though.
Original post by natninja
You can use the corrected value for the junction but do not attempt to rebalance the other flow rates during the iteration


in the second trial for the first loop , should i use positive 0.05833 or 0.01731??
Original post by wilson dang
in the second trial for the first loop , should i use positive 0.05833 or 0.01731??


Just use the final values from your previous iteration.

In fact, just use uncorrected values. It will make your life easier.
Original post by natninja
Just use the final values from your previous iteration.

In fact, just use uncorrected values. It will make your life easier.


do you mean 0.01731? which is in the first iteration after correction in the first loop
Original post by natninja
Only use corrected values for loops after the correction has been done.

It's simpler to use the uncorrected value, even if it could take an additional iteration or two though.


one more question , is it a must to change the positive value of 0.01731 to
negative 0.01731 for the first trial in the second loop ? the rule of changing positive to negative ( clockwise to counterclockwise ) is it always true for 2 adjacent loop ? or it depends on the situation ?
Original post by natninja
Just use the final values from your previous iteration.

In fact, just use uncorrected values. It will make your life easier.


which uncorrected value do you mean ?
Anyonr can answer??
Original post by natninja
Just use the final values from your previous iteration.

In fact, just use uncorrected values. It will make your life easier.

do you have any idea ?
bump
Original post by natninja
Just use the final values from your previous iteration.

In fact, just use uncorrected values. It will make your life easier.


bump

Quick Reply

Latest