Introduction

Pavement damages resulting from repeated loading of bituminous mixture layers are of great importance. In order to characterize fatigue in mixture and pavement design procedures, it is necessary to describe the behavior of bituminous mixtures under repeated stress (or strain cycles). Various laboratory fatigue tests are proposed (LPC 1977 [1], Assi 1981 [2], SHRP 1990 [3]. Eurobitumen 1995 [4]) but large scatterings are observed between the different methods (Francken & al 1995 [5]). The cyclic fatigue tests are performed either at constant stress, or strain, or dissipated energy per cycle, or even at various solicitation amplitudes.

The traditional fatigue criterion is linked to the number of cycles giving half the initial rigidity of the sample. This variation of rigidity is interpreted as a variation of the complex modulus (*E**). For example, in France, the fatigue criterion is the strain amplitude (,6) which induces half a rigidity for one million cycles, considering a constant strain amplitude two points bending cyclic test on prismatic samples. A rational approach in order to link the fatigue criteria obtained from the different monitored type cyclic tests is not yet proposed. The intrinsic character of these criteria can then be discussed. Furthermore, classification of mixtures given from laboratory fatigue tests is not always represented on the road. It appears that fatigue of bituminous mixtures is not very well modelized with the traditional approaches.

In this article, starting from an analysis of the existing procedures for fatigue simulation, we propose a new approach that isolates the pure fatigue phenomenon. This approach has been designed, at the "Département Génie

Civil et Bâtiment" of "Ecole Nationale des Travaux Publics de l\'Etat". Non homogeneity of the sample and heating effects can explain, in our sense, why the "classical" tests are not able to give directly correct information on fatigue. Let us underline that we do not deal with

Fatigue Damage 143

Figure 1. Different Stress Paths for Fatigue a)Cannot be Used for Fatigue Characterization; b), c), and d)Can be Used for Fatigue Characterization

crack propagation at a "macro level" but with the damage process induced by the application of a large number of "small" loadings. This application creates, first, damage at a "micro level" which are at the origin of the fatigue degradation of the road. As a consequence thermal cracking or reflective cracking had to be analyzed with other developments (Di Benedetto & al 1993, Jacobs 1995 [6,7,8]).

Experimental results using the new procedure for fatigue damage are presented. A fatigue law is proposed.

A prediction model with this law of a stress control test from strain control results is presented. The comparison of this model with experimental data gives a validation of the proposed coherent approach. An intrinsic fatigue parameter criterion is also given.

Classical procedure: analysis and problems

During the laboratory tests, which are used to characterize fatigue, a repeated number of cycles are applied to the sample. The large amount of cycles can be applied with or without rest periods. An illustration of the different types of solicitation used in fatigue testing is given in Figure 1.

在这篇文章中,由对当前存在的疲劳模拟现有的程序的分析开始,我们提出一种隔离纯疲劳现象的新方法。 这种方法已经被法国国立国家公共工程学院的民用建筑工程学院所构思。在我们看来，没有任何的同质性抽样 和热效应可以解释为什么"古典"测试不能直接给与疲劳特性以正确信息。 让我们强调指出，我们并不做处理

第143 疲劳损伤

图1。疲劳的不同应力路径  a)不能用于疲劳的、表征; b),c),以及d)可以用于疲劳描表征。在宏观层面上的裂纹，但是其损害过程是由于许多"小"负荷的应用所引起的。这种应用，首先造成了微观层面上的损伤，而这种损伤恰恰是道路疲劳损伤恶化的源泉。 因此热裂化和反射裂缝必须由其他发展阶段来进行分析(di Benedetto & al 1993， 雅各布斯1995 [6,7,8 ]).

在这篇文章中,由对当前存在的疲劳模拟现有的程序的分析开始,我们提出一种隔离纯疲劳现象的新方法。 这种方法已经被法国国立国家公共工程学院的民用建筑工程学院所构思。在我们看来，没有任何的同质性抽样 和热效应可以解释为什么"古典"测试不能直接给与疲劳特性以正确信息。 让我们强调指出，我们并不做处理

第143 疲劳损伤

图1。疲劳的不同应力路径  a)不能用于疲劳的、表征; b),c),以及d)可以用于疲劳描表征。在宏观层面上的裂纹，但是其损害过程是由于许多"小"负荷的应用所引起的。这种应用，首先造成了微观层面上的损伤，而这种损伤恰恰是道路疲劳损伤恶化的源泉。 因此热裂化和反射裂缝必须由其他发展阶段来进行分析(di Benedetto & al 1993， 雅各布斯1995 [6,7,8 ]).

使用疲劳损伤的新程序所得出的实验结果被提出。一条疲劳法律被提议。一种预测典型伴随着应力控制测试的法律被提出了，而这条应力测试法律是应变控制的结果。这个典型与试验数据的比较使得被提出的协调方法得以生效。 固有疲劳参数标准也被证实。

 古典程序: 分析和问题

在旨在表现疲劳特性的实验室测试期间,重复的循环数量被应用于这个样品。大量的周期可以适用无论有无休息时间。在图1中我们可以看到一张在疲劳试验过程中不同诱发类型的插图。

在” A “类诱因中， (见图1 )疲劳现象被不可逆应变的积聚而隐藏，样品的瓦解通常也发生得非常迅速。 这个解释使得在疲劳表征上进行的圆柱劈张试验无效。

晕，翻译的狗屁不是！鄙视自己一下，最近衰死了！

海鲜出来，安慰一下了！

同情.

草，1小时翻译这么多`````````

是杂人才啊`

为白白死去的几千万脑细胞

默哀

好专业的文章啊````可怜的娃娃又要再翻译一遍

沧海最近出门晃荡去啦

不能上网   呵呵

米事   偶来安慰你

唔~ 好多专业词汇，娃娃辛苦啦，休息下，吃点好吃滴犒劳下自己````

刚回长沙，累了个半死。。。。附带满身包

被蚊子咬得吧，最讨厌的就是想专业词汇，摸不着头脑，所以好乱，没准都不对，就这样吧，不想弄了，顺便说句，我生病了，感冒，真难受！