JNIOSH

Abstract of Special Research Report (SRR-No.15)

National Institute of Occupational Safety and Health, Japan

Prevention of Failure of Overhead Travelling Cranes Due to Degradation of components

Introduction

SRR-No.15-1
Yoshio KITSUNAI

: The fatigue life of overhead travelling cranes in general has been designed based on the strength of the stress range at 2 x 106 cycles for the components. Recently the operation rate of cranes has remarkably increased year by year due to economical request. For example, according to a report in a steelworks, the recent operation rate per year of overhead travelling cranes reaches about 10 times as compared to 20 years ago. As a result, some of cranes constructed in the 1960's have already reached the design life. So that the fatigue cracks are often found at highly stressed regions such as weld toe of gusset plate and of rib plate. Once the fatigue cracks are initiated, they continue to grow slowly in the early stage of the fatigue crack growth life. As long as crack growth occurs, the potential for more serious distress exists. Therefore the cracks must be repaired at an early stage before the cracks grow up to the critical crack size. Rational methods of maintenance as well as of design are required to achieve an optimum structure throughout its service life and to secure safety for workers.
    In view of the potential seriousness of cranes cracking associated with accumulation of fatigue damage and degradation of the materials, an extensive study of the problem was planned. The objectives of the study are to determine fatigue properties of welded joints used in overhead travelling cranes under service loading, to estimate how much time does it require for a small crack to grow into a serious crack and to determine the best welding procedures to repair the cracks.
    The first phase of the study includes survey of damage in overhead travelling cranes to obtain information on causes of the cracking. For the application of damage tolerance concepts, loading variables such as amplitude, frequency and sequence are important. Therefore monitoring of stresses acting on the members in an overhead travelling crane was performed under the service conditions, and rain flow method was used as a stress counting technique. The main causes of fatigue damage in cranes are related with the stress concentrations at the toe of terminating fillet welds and the welding residual stresses. To establish the basic concept for evaluating the fatigue crack growth life of the welded crane components with residual stresses under service loading, the fatigue crack growth behavior of welded joints subjected to a two-step program loading was evaluated using fracture mechanics approaches. The fatigue initiation and the propagation lives of gusset pales which tend to be the origin of fatigue cracking were examined under the program loading which was determined based on the stress measurement of the crane. The prediction of fatigue crack growth life of the gusset plate under the program loading was carried out using fracture mechanics, and the results were compared with the experimental results. To determine a minimum requirement for the quality of the repair welding, the fatigue strength of several kinds of repair weld specimens was examined under the program loading, and was compared. Based on the test results, a procedure for the repair weld for cracking in existing cranes was proposed.

Survey of Degradation for Overhead Travelling Cranes

SRR-No.15-2
Yoshio KITSUNAI and Yutaka MAEDA

: In recent years the operation rate of overhead travelling cranes has increased due to economical request, so that the damages mainly associated with fatigue have sometimes been reported near welded area such as a toe of termination of fillet weld which has been subjected to high stresses. Hence survey of damages in connection with fatigue and degradation of the material was performed using existing literatures and based on field search. This survey involves information on damage causes, locations, service term, capacity or hoisting load etc. The results of the survey may give us useful information on detection of damage and determination of inspection term for cranes.
    The damages of cranes begun to appear after three to four years service, and most of cranes have been damaged after 15 years or more service. The initiation life of the damages tends to decrease with increasing hoisting capacity of crane. This result shows that the margin of design for the cranes with large hoisting loads such as ladle crane and stripper crane is not enough as compared to the cranes with small hoisting capacity such as an overhead travelling crane with crab. The damage causes for overhead travelling cranes are almost occupied by fatigue, and others are by abrasion, corrosion etc. The locations of damage in runway girder are classified as follows; the throats of the top flange to the web or to the stiffener, the welded joints between a bottom flange and attached members such as gusset plates, stiffener terminations in web, and auxiliary lattice members and their joints. These damages in runway girders are mainly caused by the partial penetration welds and local stresses due to the concentrated wheel loads. The majority of damages in travelling crane itself was concentrated in main girder. The locations of the damage were essentially the same as those of the runway girder, that is, the weld toe of gusset plate, top flange to the stiffener connection and web at termination of stiffeners. Other damaged locations were hoist gear, travel gear, traverse gear, end carriage etc. Among the damage, 42% was related to the structure components including main girder.

Monitoring of Service Loading for Overhead Travelling Crane

SRR-No.15-3
Yutaka MAEDA, Yoshio KITSUNAI and Etsuji YOSHIHISA

: Cranes have been experienced loads or stresses of variable amplitude in random. In case of application of damage tolerance concepts for cranes, a reliable prediction must be made of number of load cycles that will propagate a crack from a certain starting size to the permissible size. The prediction of fatigue crack growth rate and growth time of a crane requires the input of relevant crack propagation data and stress history. In order to produce reliable fatigue life or crack growth data for cranes, stress spectra acting on the main girder in an overhead travelling crane was examined under service condition. The main girder of the crane used was composed of conventional pipe structure. The principal specifications of the crane were: the maximum hoisting load was 49 kN, the span was 16.3m, and the crab trolley wheel base was 3.25 m. The monitoring of the stresses was carried out using strain gauges glued on the selected components of the girder. The data obtained were recorded into a histogram recorder and a data recorder. Rain-flow method was used as a stress counting technique. The stresses acting on the girder are a function of the hoisting load, movement of the crane and frequency of operations. A 49 kN weight which corresponds to the safe working load was used during the stress measurement. A typical movement of the crane employed was as follows; the test weight was lifted by a crab trolley at an end of the girder, the trolley was traversed from the end to the opposite side of the girder, further the main girder with the crab trolley was traveled approximately 20 m on the runway girder, then turn back to the starting position tracing the original root, and the weight was lowered. A series of crane operation above mentioned was cycled 20 times during the stress measurement to make histogram of stress range acting on the selected components in the girder. As a result, the majority of the stresses monitored occupied by relatively low stress ranges below 20 MPa, which have a little influence on the fatigue damage of the girder. However, the girder was subjected relatively high stresses when the lifting or lowering was stopped abruptly by braking. Moreover, the stress of the component located in the middle of the girder reached approximately 30 MPa as the crab trolley traversed from one side to the opposite side on the main girder. The stress spectrum monitored for the main girder under service conditions was found to be expressed by the Weibull distribution.

Estimation of Fatigue Crack Grouwth Behavior of HT80 Steel Weldments under Variable-Amplitude Loading

SRR-No.15-4
Etsuji YOSHIHISA and Yoshio KITSUNAI

: Many welded joints exist in structural members of overhead travelling cranes. Since actual service loads for the cranes are generally random, these welded joints are under complex loadings. In the evaluation of fatigue strength of these joints, it is essential to clarify the effect of such complexity in load, which is added to the effect of welding-induced residual stress. Recently, damage tolerant design method is adopted in designing certain kinds of machines. In these machines, prompt relevant action based on the knowledge of fatigue crack growth is necessary on cracks detected by inspection in service and in fabrication. As it is considered that this design method will be adopted for many kinds of machines which include cranes, the estimation of fatigue crack growth behaviors will be a matter of great importance.
    As a step toward fatigue crack growth tests under random loadings, the fatigue tests under constant amplitude and two-step blocked loadings were carried out on HT80 steel butt-welded joints. The influence of residual stress and variable-amplitude loading on fatigue crack growth rate, da /dN, was evaluated by utilizing linear fracture mechanics.
    The main results obtained in this study are summarized as follows:
  (1) The crack growth rate, da /dN, in the welds are dominated by residual stress and the total loading cycles in a block, (N H + N L), and the ratio of loading cycles in the block, N H/N L, have little influence on the growth rate, where N H and N L are number of cycles for large and small amplitude loadings, respectively.
  (2) The crack opening stress intensity factor, K op, in the welds under two-step blocked loading remains in the constant level through each block and is governed by the maximum stress intensity factor, K Hmax, in the block.
  (3) The effective stress intensity factor range, ΔK Rem, can be calculated from the stress intensity factor range, K irmax - K op, where K irmax is the maximum stress intensity factor in which the residual stress is taken into account at each step in a block. And the average crack growth rate for a loading block in the weld, (da /dN )B, are correlated with ΔK Rem in the same manner which is used for the estimation of crack growth rate in the materials without residual stress under constant amplitude loading.

Fatigue Strength and Life Prediction for Gusset Welded Joints

SRR-No.15-5
Yoshio KITSUNAI, Yutaka MAEDA and Etsuji YOSHIHISA

: Weld toe in the gusset plate attached with chord members in overhead travelling crane is often an origin of fatigue cracks. Reliable data on fatigue strength and fatigue crack growth behavior of gusset plate are required for assessment of the integrity or determining the inspection period of cranes. The present study is focused on the following issues to assure the safety of the bridge girders in overhead travelling cranes, (1) determining the fatigue strength of gusset plates which tend to be an origin of cracking, and (2) predicting the life of fatigue crack growth for the gusset plates.
    The material used to fabricate the specimens is JIS SM490A steel plate with a thickness of 8mm. Gusset plate specimens with two different sizes were fabricated by welding. A gusset plate in the specimen was attached to one side of the main plate by partial penetration fillet welding with coated electrodes of 4 mm dia. Prior to fatigue testing, stress concentration factors of the weld toes of the gusset specimens were examined using an infrared stress analysis system. Welding residual stresses in the specimens were measured using a X-ray diffraction method.
    Fatigue test under the constant loading and the program loading which has been determined based on the stress measurement of a crane girder was carried out using two servo-controlled hydraulic testing machines attached with computers. The program loading applied the fatigue test was composed of 5 steps in a block and the total elapsed stress cycles in one block were 100 cycles. Stress ratio which is defined as a ratio of minimum stress to maximum stress in loading during the fatigue test was ranged between 0.05 to 0.2, and the frequency was kept in the range between 0.5 to 10 Hz in a sinusoidal wave form.
    It is found that the stress concentration factor of the gusset welded specimens used in this study takes around 2.4. The fatigue strength of the gusset welded specimens under the program loading determined based on monitoring of stresses acting on components of an overhead travelling crane has roughly 90MPa regardless of specimen size. The fatigue life of the gusset welded specimens is dominated by crack initiation rather than crack propagation. The crack growth rate of the gusset welded specimens under the program loading is correlated with an effective stress intensity factor range, ΔK Rem, estimated based on linear accumulation of the stress intensity factor range which takes into account residual stress at each step in a block. The prediction of fatigue propagation life using ΔK Rem agrees with experimental result in the range of error within 15 %, when the crack length is less than 20mm.

Fatigue Strength of Repaired Weldments

SRR-No.15-6
Yoshio KITSUNAI, Yutaka MAEDA and Etsuji YOSHIHISA

: Overhead travelling crane and their relative components are often experienced fatigue cracking. When crack is found in a structural member of a crane, welding is often employed to repair the crack because of its compatible with easy handling. The repair and strengthening of the component in the cranes occasionally oblige us to work at small place or high location under insufficient lighting, so that welding defects are sometimes introduced during the operation. Moreover, inadequate procedure for the repair weld acts to decrease markedly fatigue life of the repaired components. Therefore, a minimum requirement for the quality of the welding is necessary. However the influence of repair weld method on the fatigue strength of structural members was not fully understood. More research is needed to evaluate the effectiveness of the repair weld.
    The objectives of the study are to determine the relative effectiveness of various repair schemes for components of cranes and to determine the best welding procedure to repair the cracks. Since the fatigue strength of repair weld specimens was evaluated and the fatigue crack growth behavior of the repair weld specimen was estimated by applying the fracture mechanics concept.
    The gusset plate specimen and the repair welded gusset plate specimen were prepared. Moreover, four different types of repair weld specimens were also fabricated. These repair weld specimens were: the specimens with a crack passing through the thickness were repaired on one-surface or both surfaces of the specimen using welding, and were repaired by cover plate with side fillet welds. The material used to fabricate the specimens is JIS SB400 steel plate with a thickness of 8 mm The yield stress and the tensile strength of the plate are 333 MPa and 461 MPa, respectively. All welds were made manually by covered electrodes of 4 mm diameter corresponding to JIS D4326. Stress concentration factors for the gusset plate specimen and repair weld specimens were examined by infrared stress analysis system. The residual stresses in the specimens were measured by a X-ray diffraction method. Fatigue test under the program loading which was determined based on the stress monitoring of a crane girder was carried out using a servo-controlled hydraulic testing machine with capacity of 196kN. The program loading applied for the fatigue test is composed of 5 steps in a block and the total elapsed stress cycles in a block are 100 cycles. Stress ratio was 0.1 and frequency was 5 Hz in a sinusoidal wave form.
    It is found that the fatigue strength of the repair welded gusset plate specimen is somewhat higher than that of as-welded one because of release of residual stress arising from reheating during the repair weld and reduction of stress concentration of weld toe which is an origin of the fatigue crack due to the repair weld. The fatigue strength of the repair weld specimens with partial-penetration of welds decreases with increasing the initial crack length. In case of the one-surface repaired specimen with partial penetration of weld metal, it is difficult to obtain enough fatigue strength, while fatigue strength of the specimen repaired both surfaces by welding is satisfied the requirement of the design code of crane. In case of the cover plate repair specimen, the fatigue strength reduces significantly due to high stress concentration at the side fillet weld. Hence the cover plate is an inadequate repair method, and is inapplicable to cranes subjected to cyclic loading. From this experiment and some reports on repair weld, the following repair method may be recommended: removing the crack from the component by machining or air-arc gouging and making a groove, filling the groove completely from both surfaces of the component by welding, and finally finishing by grinding to reduce the stress concentration.

Concluding Remarks

SRR-No.15-7
Yoshio KITSUNAI

: In this study, survey of failure of cranes, monitoring of stress spectrum acting on components in a girder of a crane, fatigue crack initiation and propagation lives of welded joints, estimation of fatigue crack growth life, and weld repair method were examined to prevent failure of overhead travelling cranes due to damages associated with fatigue and degradation of materials. The results are summarized as follows:
  (1) Fatigue occupies almost fifty percent of damage causes of cranes and others are abrasion, corrosion etc. The damages of crane initiate to appear after three to four years service, and most of the cranes have been damaged after fifteen years service. The initiation life of the damages tends to decrease with increasing hoisting capacity of crane.
  (2) The location of damages in runway girder is concentrated at highly stressed region such as the throat of the top flange to the web or to the stiffener, the welded joints between a bottom flange and attached members and the terminations of stiffeners. The damage locations of the main girder are essentially the same as those of the runway girder.
  (3) As a result of stress monitoring of a main girder under service, the majority of the stresses is occupied by relatively low stress ranges below 20 MPa. However, relatively high stresses acted on the main girder when the lowering was stopped abruptly by braking. Moreover, the component located in the middle of the main girder was subjected to the stresses reaching around 30 MPa, when the crab trolley traversed from one side to the opposite side on the main girder.
  (4) The stress spectrum monitored for the main girder under service can be approximately expressed by Weibull distribution.
  (5) The crack opening stress intensity factor, K op, in the welded joints subjected to a two-step program loading is found to be almost the same in each block and is governed by the maximum stress intensity factor in the block.
  (6) The fatigue crack growth rate, da /dN of welded joint subjected to the two-step program loading is correlated with an effective stress intensity factor range estimated based on linear accumulation of the stress intensity factor range which is defined as K iRmax - K op, where K iRmax is the maximum stress intensity factor which takes into account the residual stress at each step in a block.
  (7) The stress concentration factor of the gusset welded specimens tested takes around 2.4. The fatigue strength of the gusset welded specimens under the program loading determined based on monitoring of stresses acting on components of an overhead travelling crane has roughly 90 MPa regardless of specimen size. The fatigue life of the gusset welded specimens is dominated by the crack initiation rather than the crack propagation life.
  (8) The da /dN of the gusset welded specimens under the program loading is correlated with an effective stress intensity factor range, ΔK Rem, estimated based on linear accumulation of the stress intensity factor range which takes into account residual stress at each step in a block.
  (9) The prediction of fatigue propagation life of the gusset welded joints using ΔK Rem agrees with the experimental results in the range of error within 15%, when the crack length is less than 20 mm.
  (10) The fatigue strength of the repair welded gusset plate specimens is somewhat higher than that of as-welded ones because of release of residual stress arising from reheating during the repair weld and reduction of stress concentration of the weld toe.
  (11) The fatigue strength of the repair weld specimens with partial-penetration of weld metal decreases with increasing the initial crack length. In case of the one-surface repaired specimens with partial penetration of weld metal, it is difficult to obtain the enough fatigue strength, while the both surfaces repaired specimens satisfies the requirement of the design code of overhead travelling crane.
  (12) In case of cover plate repair specimens, the fatigue strength markedly reduces due to high stress concentration at the side fillet weld. Therefore the cover plate is an inadequate repair method, and is inapplicable to cranes subjected to cyclic loadings.
  (13) The following repair method for crane members may be recommended: removing the crack from the damaged component by machining or air-arc gouging and making a groove, filling the groove completely from both surfaces of the component by welding, and finally finishing by grinding to reduce the stress concentration.

Other Publications

Kiyose District Map

Umezono 1-4-6,Kiyose,
Tokyo 204-0024 Japan
TEL: +81-42-491-4512
FAX: +81-42-491-7846

Noborito District Map

Nagao 6-21-1, Tama-Ku,
Kawasaki 214-8585 Japan
TEL: +81-44-865-6111
FAX: +81-44-865-6124