Unsupported Tube Length

[DK44]

I am interested to know as to :
1. How to calculate unsupported Tube length of a Heat Exchanger Tube for comparing with TEMA Unsupported lengths.
2. How much maximum deflection of Tube is allowed for the unsupported length.

 

[MJCronin]

In response to your questions:

1) I would use the methodology of ASME B31.1 (Come up with either an allowable stress or max deflection and use a simply supported beam method

2) Be aware that TEMA HX unsupported tube lengths consider the effects of tube vibration as well as "imperfect" (loose) support conditions at tube support plates. Your question "What is the maximum" depends on many factors ...

MJCronin
Sr. Process Engineer

 

[SnTMan]

DK44, 1) pretty much what is sounds like. For straight tube portion usually twice the baffle pitch, see RCB-4.4. For U-bend usupported length see RCB-4.54. For fixed tubesheet exchangers the end spaces (at the tubesheets) may be a consideration. See A.24, definition for "k". See also Part UHX-13.2, definition for "k".

2) I am not aware of any tabulated limits on deflection. For mechanical loadings you want to avoid deflection causing permanent deformation of the tubes. Tube vibrations due to fluid flow places other limits, see for example TEMA on flow induced vibration, V.1 and following. You definitely do not want the tubes contacting adjacent tubes in any way.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[EdStainless]

This usually involved calculations of vibration driven by cross flow turbulence. In some design codes they presume that the tubes are vibrating and then calculate support to prevent clashing.
In large HX I have seen many die because of excess vibration, it is a common problem.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[DK44]

Thank you expefrts for your responses.
1. My question 1 is seeking a method to calculate max unsupported length of the Tube for comparison with TEMA
I am thinking like:
i) Consider TEMA unsupported length for a particular Tube MOC and size.
ii) Consider the Tube is simply supported over the unsupported length.
iii) Consider UDL with Tube metal weight + Fluid in Tubes with in the unsupported length.
iv) Taking the properties MOE , Moment of Inertia, calculate the max deflection from the Beam formula for UDL as (5/384)*(w*l^4/E*I).
If this method OK,

2. My question 2 is how much deflection is allowed.

 

[EdStainless]

There are a bunch of accepted methods for calculating this (Coit and MacDuff-Fegler come to mind).
So for max defection you have two things to consider, clashing and fatigue.
Typically to reduce the risk of clashing people keep the deflection to 1/3 of the ligament.
Determining what load to use for the deflection calculation is the trick. If it is a fluid that is on the shell side then the momentum imparted and turbulence caused by cross flow will be factors. As I recall these forces end up being 3rd or 4th power with flow velocity.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[SnTMan]

DK44, baffle type, spacing, need for tube supports, etc is normally determined by thermal and hydraulic considerations. Heat transfer and flow chacteristics within the shell. Determined during rating of the exchanger. These things determine the unsupported tube length. It may also be affected by any fixed tubesheet calculations of tube compressive stresses.

If I undestand your proposal it is to calculate the maximum span based on gravity loads. No offense, but a waste of time in my opinion. Other considerations dominate.

For limits, refer to Eds' posts.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

As say SnTMan "Other considerations dominate" : agreed
Think on cleaning and replacement of tubes which are affected by the baffle spacing.
It is difficult to install a new tube when the distance between baffles is large (fixed tubeshet).

Regards

 

[r6155]

Practical manufacturer: place a tube on 2 supports and measure the deflection and distance between supports as you wish.

Regards

 

[DK44]

1. Yes r6155. This what is in my mind. Hence the beam formula sighted by me can be used to calculate deflection I suppose.
Is the UDL composed of self weight of Tube in the span + Fluid (say Water)content in the tubes?
Is permissible deflection 1/3 Ligament as Edstainless indicated?

2. Request experts to advise as to what is the basis of TEMA for fixing unsupported Lengths for various Tube MOCs

 

[SnTMan]

DK44, 2. Don't know for sure, but I'd guess to avoid excessive deflection.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[EdStainless]

DK44, the gravity load is likely only 10% of the lateral load on the tubes in service. So why care about it?
Who are you getting the rest of your loads?


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[DK44]

TEMA Table for max unsupported length says at foot note that the indicated max unsupported lengths do not consider potential flow induced vibration problems. Then what is the criteria for the tabled unsupported lengths? What deflection is considered for such unsupported lengths? Can any body tell?

 

[SnTMan]

You might submit an Inquiry to TEMA

Edit: Added link

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[EdStainless]

DK, This is why people use other methods for determining support spacing.
The TEMA just gives the minimum (well in this case maximum) that can be used.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy

 

[r6155]

DK44. All codes and standards are based upon sound engineering principles, research, and field experience in the manufacture, design, inspection, installation, testing, use, etc., etc.

As say SnTMan: ".......waste time....."

Regards