On the effect of throughout layer thickness variation on properties of additively manufactured cellular titanium structures

Abstract

Over the past two decades, additive manufacturing has opened a new window of opportunities in fabricating complex porous matrix structures such as cellular solids. Several factors including design, material and process parameters can selectively be varied to tailor the porous properties of products based on the intended application. This article addresses the effect of variable throughout layer thickness configuration in the binder-jet additive manufacturing of titanium structures for orthopedic applications. Two layer thicknesses of 80 and 150 μm are selectively controlled inside of each titanium sample with four different configurations. Several studies were performed, including shrinkage analysis, porosity measurements, and mechanical compression tests to quantify the effect of layer thickness on part quality and mechanical properties. The results of the porosity measurement revealed that there is about 5% variation among the samples with different layer thickness configuration. Bulk porosity values obtained from micro computed tomography (μCT) scan data placed the bulk porosity of the samples combining more than one layer thickness, in between of the results for control specimens, which were manufactured by applying a single layer thickness throughout the samples. Mechanical properties did not show any significant variation, which is attributed to the low range of the porosity deviation (less than 5%). The highest Young’s modulus of 3.50 ± 0.4 GPa and yield stress of 175 ± 25 MPa were obtained from analysis of the data achieved from the compression test.

Introduction

Porous cellular structures have become increasingly relevant in different fields of study due to their selectively tailored properties, such as controlled stiffness, thermal conductivity, fluid permeability, acoustic and magnetic properties, impact energy absorption, and electrical isolation, as well as lightweight design [1], [2], [3]. Controlling the cell size and distribution throughout the structure can demonstrate the variation in each of these properties within the sample and classify it as a functionally graded structure (FGS) [4]. The footprint of FGS configurations can be found in nature, with human bone as one of the well-known examples [5], [6]. The in-depth knowledge of the physiological micro- and macro- structure of bone, in combination with a variety of manufacturing methodologies and tools, has generated promising results in the design and fabrication of bone implants [7].

For bone substitutes or implants, specific manufacturing criteria need to be addressed to successfully implant and stabilize the device in a patient’s body [7], [8]. Such criteria include the bio-compatibility of the material, bulk and localized mechanical properties, as well as design, featuring the anatomy of the bone and its internal structure such as pore morphology. The last factor is essential for allowing body fluid permeation throughout the implants to enhance the bone growth and implant-bone fixation [6], [9]. In addition, porosity directly affects the mechanical properties of bone substitutes and control of this parameter can simulate the appropriate bio-mechanical response with the surrounding tissue [10], [11].

The development of additive manufacturing (AM) processes [12], (also known as 3D printing), combined with image processing techniques, has created a substantial opportunity for tailoring the design of implants according to each individual patient’s anatomy [13], [14], [15]. The binder-jet AM process, one of the AM classes, constructs a 3D object in layer by layer fashion via spreading a thin layer of powder and injecting a binder to provide structural integrity [16], [17], [18]. A heat treatment cycle is often required as a post processing step to achieve the final product. Several parameters, such as powder size, shape and distribution, binder material and saturation level, layer thickness and sintering protocol have been mentioned in the literature as influential factors on the properties of the samples [19], [20], [21], [22], [23].

The layer thickness, by affecting the applied compaction force on each layer, can tailor the porosity and mechanical properties of the samples. Several studies have been presented by researchers to evaluate the effect of this parameters via employing different methods of AM [24], [25], [26], [27], [28], [29], [30], [31], [32]. According to the proposed results, decreasing the layer thickness, by increasing the compaction force applied on each layer, has a direct effect on enhancing the density and mechanical strength of samples. In these studies, only one layer thickness has been applied throughout the entire part during the manufacturing process. It can be due to the lack of ability to selectively change the layer thickness in real-time and accordingly adapt the slicing of the 3D objects in the commercially available 3D printer software.

The aim of this study is to manufacture functionally graded porous Ti structures by incorporating a variable layer thickness throughout the sample at pre-defined locations via a binder jetting AM process. The 3D printer system in this work has been developed in at the University of Waterloo [Waterloo, Canada], which provide the possibility incorporating different processes parameters as compare to the commercially available binder jetting system. The detail of this system specification has been described elsewhere [33]. In this study, four categories of the samples were designed. The first two were AM- made with a single layer thickness throughout to act as a control reference (A and B); and in the second two categories, the layer thickness was varied from high to low to high (150 μm/80 μm/150 μm) (C) and from low to high to low (80 μm/150 μm/80 μm) (D) in each batch, respectively. Sample C and D were designed with a symmetrical distribution of layer thickness with similar weight (0.5) to investigate the effect of layer thickness arrangement on mechanical properties of the specimens. To select the minimum layer thickness (80 μm), one step calibration was performed to adjust the roller speed to achieve a reasonable print quality by avoiding part distortion and cracking between two consecutive layers and considering the range of the powder particle size. Then the maximum layer thickness was selected approximately two times greater than the minimum (150 μm) to provide the reasonable variation when samples were studied.

A set of analyses were performed to evaluate the relation between the layer thickness and properties of the samples, including shrinkage analysis, porosity measurement, mechanical strength, stiffness and SEM analysis. In addition, to study the internal distribution of pores based on the layer thickness, the data from μCT scan of one sample from each group was obtained and analyzed with an image processing method to provide more clarified understanding of the internal structure of the samples. The bulk porosity, pore size and distribution as well as sinter neck size and distribution were obtained using the μCT analysis.