ML20235F681

From kanterella
Jump to navigation Jump to search
Rev 0 to IM-P-013, Justification for Use of Rigid Clamp Stiffnesses
ML20235F681
Person / Time
Site: Comanche Peak  Luminant icon.png
Issue date: 09/11/1987
From:
ABB IMPELL CORP. (FORMERLY IMPELL CORP.)
To:
Shared Package
ML20235F613 List:
References
IM-P-013, IM-P-013-R00, IM-P-13, IM-P-13-R, NUDOCS 8709290277
Download: ML20235F681 (5)
v • d • e


Text

_ _ _ _ - _ _ . - _ _

J CLARIFICATION OF IMPELL CTH DESIGN VERIFICATION CRITERIA /HETHODS FOR RESOLUTION OF CYGNA ISSUES 1

l Justification For Use of Rigid Clamp Stiffnesses l

l Prepared for:

Texas Utilities Electric Company Prepared by:

Impell Corporation 0210-040/041 l

IM-P-013 Revision 0 Prepared by: Afd- 1 l # f["7 Approved by: bb 7['/[b7 l

/

8709290277 870923 PDR A

ADOCK 05000445 PDR Page 1 of 4 1

.-_--._---_---___j

IM-P-013-

' ISSUE: Cygna has requested Impe11 to provide.a summary paper describing the change.in modelling procedure from the use of flexible to rigid stiffnesses for tray clamps.

BACKGROUND: For cable tray hanger' design' verification, tray clamps were originally modelled with-flexible stiffnesses [1]. These were static stiffness values derived analytically [6-& 7] by assuming a load r transmittal location between the tray. and the clamp and applying a unit force / moment at that point. The use of flexible stiffnesses resulted in predicting-unrealistic vertical modes which amplified the system seismic response. To develop-cable tray system models which exhibited more realistic dynamic properties and reduce this overconservatism, the use of more realistic rigid clamp stiffnesses was investigated. Indeed, test correlations [2 & 3]

showed the rigid stiffnesses for five degrees of freedom yielded more reasonable vertical modes and predicted displacements. Hence, rigid stiffnesses for these five clamp local degrees of freedom (Kx, Ky, Kz, Kxx, Kyy as shown in Figure 1) were incorporated into PI-11 [4] as a refined modelling procedure.

DISCUSSION: The change from. flexible to rigid stiffnesses for tray clamps is justified for the following reasons:

1) The original " flexible" stiffnesses were relatively rigid for all local degress of freedom except Kx (vertical axis translation) and Kzz (transverse axis rotation). Accordingly, the dynamic response of cable tray systems for these directions did not change significantly when the modified " rigid" stiffnesses were used.

This is demonstrated in Impell Calculation M-28[5], by comparing the minimum clamp stiffness for all support types (longitudinal and transverse) to the support stiffness of a typical cable tray hanger. This comparison showed the clamp stiffnesses to be significantly larger relative to the support stiffnesses for all directions except Kx and Kzz. Since the clamp and support assembly act as springs in series, any increase in the already large clamp stiffness will not significantly change the overall or effective stiffness of the assembly.

Therefore, the natural frequencies of vibration and subsequent dynamic response of the cable tray system 'for these directions remain virtually unchanged.

i 1

Page 2 of 4 i i

i e

?

IM-P-013 DISCUSSION: 2) The use of a rigid vertical stiffness K xcan be justified analytically and was also confirmed through test data. The " flexible" stiffness value was originally calculated by applying a vertical uplift load at a lower bound stiffness location along the clamp. A rigid vertical stiffness would more appropriately model the

. downward transfer of load directly from the tray to the support. This is the load transfer path during the majority of the earthquake duration when seismic Jplif t does 'not overcome gravity.

The flexible vertical clamp stiffness K x was overconservative and predicted unrealistic vertical modes. These modes were characterized by significant modal amplitude over several-supports, very high mass participation, and frequencies at or near the broadened spectral peak. The change in the vertical stiffness from the " flexible" to the " rigid" values resulted in a more realistic dynamic properties and more reasonable response predictions.

This is confirmed in the system dynamic test correlation done in [2] and (3). These correlations showed that when the flexible vertical clamp stiffnesses (Kxmin - 3.60 K/in) were used in the analytical models, the predicted vertical displacements and accelerations were grossly conservative compared to test data. Some margins of overprediction were nearly 2500%.

Furthermore, predicted relative vertical displacements at the support locations were very large (i.e., approximately 0.50 in.).

Displacements of this magnitude would result in gross plastic clamp deformation. No such behavior was observed during the system testing.

Contrasting 1y, when rigid clamp stiffnesses (K x

- 104 K/in) were substituted, the predicted displacements and accelerations were more reasonable with overprediction margins consistent with all other degrees of freedom. A comparison of vertical modes and a further discussion of the test correlation with flexible and rigid clamps is also included in Impell Calculation M-28 [5].

3) The use of rigid clamp stiffnesses facilitated l- the design verification effort. Using a constant i rigid 4 K/in transnational and 10glamp stiffnoss K in/ rad (10 is more convenient rotational) than using various flexible clamp stiffnesses i which depend on tray size, clamp type and degree of freedom. By using a rigid stiffness for all degress of freedom except Kzz (which remained unchanged) the analyst could facilitate the design verification effort. In addition to Page 3 of 4 i

K ZM-P-013.

e convenience, rigid stiffnesses reduced the margins of overprediction (while still maintaining adequate conservatism) and hence were used as a possible means of eliminating unnecessary modifications.

~

CONCLUSION:

The original flexible stiffnesses were already relatively rigid except for Kx and Kzz-Therefore any increase in the clamp stiffness to a more rigid value produced an insignificant effect on system response.

The use of a more rigid vertical clamp stiffness Kx more appropriately models the downward transfer of load directly from the tray to the

-support tier.

i The increase in vertical clamp stiffness Kx to a more rigid value eliminated unrealistic vertical modes and provided more reasonable-response prediction while still maintaining adequate levels of conservatism. This was demonstrated by comparing analytical to test  :

results.

Rigid clamp stiffness values were subsequently implemented to facilitate. design verification and provided a means of reducing unnecessary modi fica tions .  !

REFERENCES:

1. Impe11 Project Instruction PI-02, " Dynamic  !

Ana'iysis of Cable Tray Systems", Rev. 5.

i

2. Impe11 Calculation TC6-PT1 " Post Test Analysis Test Case 6," Rev. 2.
3. Impe11 Calculation TC7-PT1 " Post Test Analysis Test Case 7," Rev. 1.
4. Impe11 Project Instruction PI-11. " Cable Tray l System Analysis and Qualification Closecut", Rev. i 2.

l

5. Impe11 Calculation M-23 " Justification of Clip '

Modelling Procedure", Rev. 3.

6. Impe11 Calculation M-10 " Cable Tray Clip Angle Sti f fnes s, Rev. 2. {

l

7. Impe11 Calculation M-19 " Clip Stiffness Production Value", Rev. 2.

Page 4 of 4 i

'9 e

X = TRAY VERTICAL DIRECTION-

. Y = TRAY LONGITUDINAL DIRECTION Z = TRAY TRANSVERSE DIRECTION Z _

I ,

/ l 9

Y X l

f FIGURE I l TRAY CLAMP COORDINATES

_

Retrieved from "https://kanterella.com/w/index.php?title=ML20235F681&oldid=3191348"