ML17300A586

Evaluation of Block Masonry Walls at Palo Verde Nuclear Generating Station, Technical Rept

ML17300A586
Person / Time
Site:	Palo Verde
Issue date:	09/30/1986
From:	Hamid A DREXEL UNIV., PHILADELPHIA, PA
To:
Shared Package
ML17300A584	List: ML17300A586
References
NUDOCS 8610140291
Download: ML17300A586 (14)

Text

~ ~

App d;~ p,

~ I Technical Report on EVALUATIONOF BLOCK MASONRY'WALLS AT PALO VERDE NUCLEAR GENERATING STATION submitted to Nr. Steve Triolo Engineering Department Franklin Research Center Philadelphia, PA Dr. Ahmad A. Hamid, Ph.D., P.E.

Associate Professor of Civil Engineering Drexel University Philadelphia, PA September )986 8610140291 86100b PDR ADOCK 05000528I P PDR

~ ~

0 f

I I

1- I HTRODUCT I ON

,Block masonry walls at Palo Verde Nuclear Generating Station (PVNGS) Un) ts 1,2 and 3 were constructed with lap splices for the vertical reinforcing steel. Splices, which were not staggered, have lengths less than that specified in the AC) 531 masonry code (1) for reinforced masonry construction. The NRC staff and consultants have expressed their concerns regarding the bond stresses at the splices and the margins of safety under SSE and OBE earthquake loads for walls at 74 ft Elevation.

The NRC staff and consultants visited the plant on March 20,1986 and inspected the masonry walls. Several meetings were conducted at the NRC to discuss different aspects of the problem. Two reports dated April 16, 1986 (2) and tune 19,1986 (3) were submitted by the licensee regarding masonry wall evaluation at PVNGS.

This report presents a review of the tune 19,1986 report regarding the technical evaluation of masonry walls at PVNGS.

2- ANALYT!CAL METHODOLOGY Time history analyses were performed by Bechtel on coupled models that included representations of both the control building structure and the masonry walls. Theso) 1-structure interact)on was considered in this study. A lumped mass model of the control building was used to develop

~ ~

t

the response spectra. A stick model of I- ft. strip of the wall was used to analyze the masonry wall at Elevation 74 ft. A single direction T-H record was used to analyze the wall using finite element method adopting a macro-analysis approach (Le.,mortar joints were not modelled). A number of assumptions was used in this analysis; an evaluation of each is presented in the following section.

3- EVALUATIONOF ASSlNPTIONS

/) S/ngl~ d/I erat/on time h/story-This represents a realistic approach since the masonry walls in question are nonloadbearing elements for which the out-of-plane behavior dominates their response.

//J Strip /deal/zat/on of'he sall- The wall behavior. is assumed to be one-way in the vertic8l direction which is a conservative and a realistic assumption because the side boundaries of the walls are free. Also, the wall pattern is a running bond and openings are adequately reinforced which assure continuity in the hor izontal direction.

/1/JPfater/a! propert/es- A conservative grout strength and average rebar location were used. Mall modulus of elasticity is assumed to be equal to l 000 f'm, where f'm is the prism compressive strength. This formu)a, which is specified by the current masonry codes (),4) highly overestimates the elastic modulus and would lead to nonconservative estimate of wall stiffness (5). This is an important factor to be considered in the evaluation since PVNGS wall response is highly

~ ~

I L'

sensitive to the calculated frequency.; see the critical range in the response spectra presented in Fig. 1.

/v) S-stage moment o/ /nert/a- It is assumed that the wall undergoes three stages of cracking:I) uncracked, 2) partially cracked where only the faceshell is cracked (I.e.,mortar debonding), and 3) fully cracked when the tensile stresses in the extreme fibers of the grout cores reach modulus of rupture of the grout. Test results (6) do not support the Bechtel assumption of 3-stage cracking model. The tests indicate that cracking of the faceshell will occur simultaneously with cracking of the grout and that grouted masonry, as a composite material, has only one cracking moment. The Bechtel approach is neither realistic nor conservative in estimating wall stiffness.

4- ANALYSIS OF RESULTS The time history analyses coupled with the 3-stage model surprisingly revealed very low bond stresses (110 psi for SSE and 80 psi for OBE at Elevation 74 ft) compared with those from previous simpified analysis presented in the April, 1986 report (2). Bectel concluded that walls at PVNGS are adequate because calculated bond stresses were below the code allowables ( 180 psi for SSE and 120 psi for OBE).

The calculated bond stresses are h/ghly sensitive to the estimated wall stiffness. The stiffness determines the wall frequency which in turn

II II

COMPARISON OF PUBLISHED AND ENVELOPED WALL SPECIFIC SPECTRA 7T DAMPING 1.0 I

I I

I I I I

IIII IIII I I I

I I I

I I

Ill III 0.8- I I

I I I I III III l I I I I

-I- I- I- I-IA 1

III 4.

T 0

I I I I I I I

- 7 "I I III I 1 I I I I 0.6-I F' I I J

I 2:

O I I I I III 1 NOTE: Wall s P ecific casos Includ es 1.5Es, I.OEs, .67Es I- L I LLLI I I I II I (X I I I I II 0.4 1 I I I 17 4J O I I

I g'f IIII I O I T

0.2 I

I I-I I

I I

I IIII I I I I I- l- I-Id I I I I ] III PUBLISHED (ocakd 4 .26)

~ ENVELOPED WALL. SPECIFIC 0.0

.01 ~ 10 PERIOD (SEC) 10.

range of wall response Flg, l- Floor Spectra at Elevation 74 Ft.

i ~

li

'etermines the induced loads and the resulted stresses the wall will experience during an earthquake. The small change in wall'frequency results in a large change in acceleration due to the fact that the period of PVNGS walls in question falls in the steep portion of the response spectra curve; see Fig. l.

A proper design method should conservatively account for the sensitivity of the calculated frequency and the inherent variability in estimating the modulus of elasticity and the effective moment of inertia of such a brittle material as masonry and concrete (7).

5- CONCLUS ION Based on the review of the information submitted in tune,1986 repor t (3) and discussions of concerns presented above,it.is concluded that Bechtel design methodology of PVNGS masonry walls regarding the calculation of wall stiffness is not justified. This approach could lead to nonconservative results for bond stresses fn lap splices at Elevation 74 ft.

Therefore, it is concluded that Bechtel analytical methodolgy presented in June 19B6 report is not acceptable.

7- REFERENCES (1) American Concrete Institute 531 Code, Building Code Requirements for Concrete Masonry Structures," Detroit, Michigan, 1979.

4 II

(2) Evaluation of PVNGS Masonry Walls, from E.E. Van Brunt, Jr., Arizona Nuclear Power Project,to G.W. Knighton, NRC, Dated April 16,1986.

(3) PVNGS Masonry Walls, from E.E. Van Brunt, Jr.. Arizona Nuclear Power Project,to G.W. Knighton, NRC, dated June 19, I 986.

(4) International Conference of Building Officials, UBC Code, Chapter 24, 1985.

(5) 2iab, G,, Modulus of Elasticity of Concrete Block Masonry," M.Sc, Thesis, Depar tment of Civil Engineering, Drexel University, 1986 (6) Drysdale, R. and Hamld, A," Effect of Grouting on the Flexural Tensile Strength of Concrete Block Masonry," Proceedings of the Masonry Society Journal,Vol. 3, No.2, July-Dec. 1984.

(7)" Variability of Deflections of Simply Supported Reinforced Concrete Beams," Report by ACI Committee 435, ACI Journal, Proceedings 69, January 1972.

~

) 'I

)I I