Accepted Manuscript

Title: Personalized Hip Implants Manufacturing and Testing

Authors: A. Sorin Mihai Croitoru, B. Adrian Pacioga, C.

Stanca Comsa

PII: S0169-4332(17)30558-5
DOI: http://dx.doi.org/doi:10.1016/j.apsusc.2017.02.185
Reference: APSUSC 35294

To appear in: APSUSC

Received date: 7-10-2016

Revised date: 12-2-2017
Accepted date: 20-2-2017

Please cite this article as: A.Sorin Mihai Croitoru, B.Adrian Pacioga, C.Stanca Comsa,
Personalized Hip Implants Manufacturing and Testing, Applied Surface Science
http://dx.doi.org/10.1016/j.apsusc.2017.02.185

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
<AT>Personalized Hip Implants Manufacturing and Testing

<AU>A. Sorin Mihai Croitorua ##Email##sorin.croitoru@gmail.com##/Email##, B. Adrian

Paciogab, C. Stanca Comsab
<AFF>aPOLITEHNICA University of Bucharest, Faculty of Engineering and Technological
Systems Management, Department of Machines and Manufacturing Systems, 313 Spl.
Independenței, Sector 6, 060042-Bucharest, ROMANIA
<AFF>bNational Institute for Research and Development in Mechatronics and Measurement
Technique,
6-8 Pantelimon Road, Sector 2, Bucharest, ROMANIA

<ABS-HEAD>Highlights► Personalized implants with fenestrated design were made for

mechanical testing. ► Fenestrations reproduce trabecular structure of natural bone structure. ►
Static tests revealed an elastic behavior compared to a commercial implant. ► Endurance limit was
determined by realization of dynamic tests. ► Both fenestrated models can be safely used in the
medical practice.
<ABS-HEAD>Abstract
<ABS-P>Two models of Ti6Al4V personalized femoral stems for hip replacement have been designed and laser sintered
with different sizes of fenestrated architecture that mimics the natural structure of bone, ensuring postoperative bone
ingrowth and increasing the elasticity of the entire structure. They were tested statically and dynamically versus a
commercial femoral stem. Mechanical tests were performed in order to determine the fatigue limit using the Locati
method. The tests were conducted in a thermostatic bath (37°±1°) with the implants immersed in distilled water salted
solution 0.91%. For probe embedment poly-methyl methacrylate (PMMA) was used.
<ABS-P>The characteristic curves of the two personalized fenestrated implants reveal an elastic behaviour by their
nonlinear appearance. After dynamic tests an inverse relationship between displacements obtained in the static tests and
the fatigue limit was observed. Large fenestrations conferred the desired elasticity to the implant, but contributed to a life
service reduction.
<ABS-P>The fatigue limit for both implants was much above the minimum value specified by ISO 7602: 2010, so both
models can be safely used in the medical practice, leading to increased life service of implants.
<KWD>Keywords: 3D laser sintering; personalized implant; mechanical rigidity; fatigue limit.
<H1>1. Introduction
Experimental studies [1,2], have shown that in total hip arthroplasty, adaptation of geometric shape of the uncemented
implant stem to the inner contour of the cortical bone of the proximal femur is essential for optimizing load takeover and
for a better mechanical stability. Taking into account these studies and also the characteristic trabecular structure of the
proximal end of the femur and using the idea to reproduce (at a macroscopic scale) this natural bone structure [3], two
models of customized femoral stems for hip replacements were designed. These models had a fenestrated architecture
that mimics the natural bone structure, providing postoperative bone ingrowth and final implant fixation, enhancing the
elasticity of the entire structure, thus approaching much more to the natural elasticity of the healthy bone. During the
daily human activities in the femoral joint are induced forces that exceed three times the body weight of the patient. Due
to these forces, the internal architecture of the upper end of the femur has a trabecular structure which can be easily seen
on a section in the frontal plane. Forces acting at this level are distributed along three main trabecular patterns which
follow the direction of principal stresses, ensuring maximal strength for minimum mass of the bone as a whole.
The two models of implants were provided with two different sizes of fenestrations, the mechanical tests that followed
aiming to determine which one of the two models withstands better the loading occurring during normal use. The
personalized design was based on the CT images from a patient requiring a surgery for hip arthroplasty, their
manufacture was made using biocompatible titanium alloy powder (Ti6Al4V) by 3D additive manufacturing (laser
sintering). Implants made using additive manufacturing are presently used in some cases for patient specific anatomy to
reduce stiffness and stress-shielding effects, to obtain enhanced mechanical performances and also for new functionalities
like drug controlled delivery after implantation [4,5,6]. Additive manufacturing has opened new possibilities so complex
theoretical designs can now become reality, ensuring flexibility in manufacturing parts due to their capability to
manufacture several parts with different geometries using a single setup and with minimum material consumption.
The research is also important because although the realization of personalized implants is already known, accepted and
applied in orthopaedics, study of their mechanical strength was poor, almost not studied. However, a study regarding

1
compression-compression fatigue of selective electron beam melted cellular titanium porous structures (not an implant)
observed a lower fatigue strength than the expected value, the possible reasons for this reduced fatigue lifetime being
stress concentration and the microstructure [7].
The obtained results showed that fenestrated architecture contributed to increase the elasticity of the implant, both models
comply with the actual standards requirements regarding life service, and the fatigue limit of the model with small
fenestrations was comparable to that of commercial implant.
The crystallographic structure of sintered implants was also studied and compared to the crystallographic structure of
titanium alloy cold drawn bar in order to see if it has a greater porosity which could influence the final implant durability.
<H1>2. Materials and methods
<H2>2.1. Design and manufacture
3D design data was obtained by computed tomography (CT– SOMATOM Emotion 16, Siemens Medical Solutions,
USA), with a resolution of 512 × 512 pixels, the pixel dimension being 0.529297mm and the scanning pitch 1mm. The
images were then segmented using Mimics 17.0 program (Materialise, Belgium) [8] and the 3D model of the femur was
then exported in SolidWorks (2015 Edition, Dassault Systemes, USA) [9], a parametric design software. Design of the
femoral stems was performed in this program., which were then made of The stem shape in lateral and medial directions
is given by the inner contour of the cortical bone from the medullary cavity, ensuring a good support of the loading
forces and the stem length was established by the orthopaedic surgeon. Fenestrations with a medium size of 6x6mm and
2x2mm were made following the directions of the trabecular patterns (bundles), in order to enhance the elasticity of the
entire structure and to reduce stress-shielding. The shape of the fenestrations is given by the directions of the trochanteric
patterns in this area.
The implants were then manufactured using titanium alloy powder - Ti6Al4V by 3D additive manufacturing (laser
sintering), on EOSINT M270 Extended machine [10] (see Figure 1).
<H2>2.2. Mechanical testing
For the static and dynamic tests, implants were gripped in the position specified by the standard 7206: 2010, parts 4 and 8
[11,12], this positioning ensuring complex loading of femoral stem, simulating both compression and bending efforts and
also the torsion occurred during normal use. The gripping of test samples on the testing machines was done this way to
simulate worst-case loading, which involves the total loss of bone support in the proximal femur zone. According to
testing standard, orientation angles for the femoral straight stems without anteversion are: α = 10°±1°in the frontal plane,
β = 9°±1° in the sagittal plane and for the realization of these angles the guidance system presented in Figure 2 was used.
For sample fixation an acrylic resin was used (PMMA) that has the classic features of acrylate (elastic modulus E =
3.8GPa, Poisson's ratio = 0.3), its advantage being the satisfactory deformability (the material is resistant to dynamic
stresses) and the modulus of elasticity close to that of the natural bone [13].
<H2>2.3. Static tests
For the static tests a Hounsfield H10KT machine from Tinius Olsen [14] was used and a subroutine was conceived
whose input data were the working speed (3 mm/min) and the force at which the test stops, respectively 2700N, which is
the maximum dynamic load that occurs during walking, its size being determined based on the weight of the patient.
The samples were placed on the machine table so that the ball joint to be along the axis of the working test equipment.
For transmitting the loading force on the machine crosshead a low friction mechanism was mounted in order to minimize
the forces having a direction that does not match with the loading direction (see Figure 3).
<H2>2.4. Dynamic tests
For determining fatigue limit the Locati method was used, which takes place in two stages: one for estimation and one
for confirmation of the fatigue limit - FDL. The test starts at the minimum load level of 2300N at which the sample must
withstand 5 million cycles and then the test continues in consecutive steps on the same sample, the loading force
increasing by 500N after every 1 million cycles until the sample cracks or until the maximum load reaches to 4800N.
Based on the implementation of these steps the fatigue limit can be determined by using the following equation (see also
Figure 4):
FDL=F0,L-1+[(F0,L-F0,L-1)x(nL/Δn)] (1)
where F0,L is the maximum load at the level where the specimen fails in the Locati test, nL is the number of cycles to
failure at the load level F0,L and Δn is the number of cycles at one level in the Locati test (Δn=1,000,000 cycles).
If FDL> 3300N two additional tests are needed to confirm this limit. The two supplementary samples must withstand 5
million cycles without breaking, the loading force used for these tests being:
FD=0,95xFDL (2)
If 3300N>FDL>2300N, six additional tests are needed to confirm this fatigue limit. If one of the samples tested for
confirmation of the fatigue limit breaks, then the testing is restarted using a F DL value decreased by 5%. Locati test
parameters are shown in Table 1 and the progress of the fatigue test method is illustrated in Figure 4.
Fatigue tests are recommended to be carried out in an physiological and biochemical environment similar to the human
body, which makes possible the occurrence of fatigue corrosion and which may lead to lower the fatigue limit with up to

2
30% [15]. Therefore, the implants were immersed in a saline solution of distilled water with a concentration of 0.91%
(9.1g NaCl in 1000ml distilled water), and its temperature was maintained at 37°±1° C (see Figure 5).
<H2>2.5. The study of the crystallographic structure
Crystallographic structure of sintered titanium compared to a titanium rolled bar was studied using a metallographic
microscope Nikon Eclipse MA 100 [16], equipped with a control software NIS-Elements F Package [17], system that
ensured the automated acquisition of image and allowed to obtain an uniform image with stable temperature colour,
regardless the changing of the ambient light.
<H1>3. Results
The static testing of the three models showed that the biggest displacements occur in the implant with large fenestrations
due to its increased elasticity, elasticity which is conferred by the architecture that mimics the natural structure of the
proximal area of the femur. The obtained displacement values are presented in Table 2.
After static tests the characteristic curves (force vs. displacement) were obtained, as shown in Figure 6 for sample I of
each implant model.
Locati test results for dynamic tests carried out on the three models of implants are shown in Table 3, and the conduct
of the test is shown graphically in Figure 7 for the personalized implants.
Based on the maximum force and the number of cycles at which the sample resisted, according to equations (1) and (2)
the estimated fatigue limit FDL and fatigue limit FD were calculated. For all three samples estimated fatigue limits F DL
greater than 3,300N resulted.
For the commercial prosthesis the progress of the test has not been graphically illustrated, but taking into account the
values obtained for the fatigue limit F DL, it is almost identical to that of the prosthesis with small fenestrations except that
the number of cycles withstood by the implant at the force level of 5,300N was 63,428, which corresponds to an
estimated fatigue limit FDL = 4,832N.
The study of the crystallographic structure conducted to the images presented in figure 8, representing the microstructure
of titanium alloy:
- Figure 8,a Ti-6Al-4V rolled bar. This material is used by Tehnomed Impex Co S.A., a Romanian manufacturer for
dental implants and other implantable devices;
- Figure 8,b presents the microstructure of a sample of EOS Ti64, obtained by laser additive manufacturing of
metallic powder.
<H1>4. Discussions
<H2>4.1. Static tests
In the characteristic curves (see Fig. 6) slight discontinuities can be observed, which are due to micro displacements that
took place between acrylate polymer and the tapered surface of the gripping device (wedging on cone). The characteristic
curves of the two personalized implants with fenestrations reveal an elastic behaviour by their nonlinear appearance. In
the implant with large fenestrations the curve has a pronounced nonlinear elastic behaviour similar to the rubber, showing
larger strains at smaller forces, strains that decreased with the increase of the loading. By contrast with these two models
for the commercial implant, variation force versus strain is almost linear, the curve being similar in terms of appearance
to that of a steel subjected to compressive zone I of the characteristic curve - the elastic strain zone.
Although at first glance the value of maximum strain for large fenestrations model seems high, considering the
provisions of the standard for fatigue test that stipulates a maximum allowable displacement of 5mm when loading the
implant, that value is acceptable. It is noted that commercial prosthesis is on the next position in terms of displacement
value, on the last position being the implant with small fenestrations for which the largest displacement was 0.981mm
and average displacement 0.931mm.
The authors did not find in specialized literature similar tests but in an article regarding computer simulation of femoral
stem [19] are presented total displacements up to 6.57mm, so higher than those determined experimentally by the authors
of this paper.
It should also be pointed out in this context that some influence on the appearance of characteristic curves and values
obtained for total displacement has, clearly, also the material used for embedding the specimen, which has its own
elasticity whose influence cannot be quantified by these mechanical tests.
<H2>4.2. Dynamic tests
The conclusions of dynamic tests can be summarized as follows:
- Fatigue strength of the implant depends on the geometry of the femoral stem and the length of its neck;
- The safety margin (an indicator in predicting the longevity of the implant) which is the difference between the fatigue
limit and the maximum dynamic load (2700N) of the implant is: 1,890N for commercial implant; 676N if the implant s
with large fenestrations; 1,982N in the case of the implant with small fenestrations.
This means that the commercial implant and small fenestrated implant will be more protected in case of accidental
overloads due to unusual activity of the patient or other unknown factors;
- There is an inverse relationship between strains obtained in the static tests and the fatigue limit. Thus, for large
fenestrations model for which resulted the most significant displacement (s = 2.28mm), the fatigue limit was the lowest;
- Large fenestrations conferred the desired elasticity to the implant, but contributed to reduction of its life service;

3
- Since all samples were made of the same material (Ti-6Al-4V), authors could not assess the influence of the base
material on fatigue limit;
- Fatigue test was conducted according to actual international standards, which stipulates sample testing in the worst-case
scenario, respectively the loss of bone contact in the proximal zone of the femoral stem. This leads to the annihilation of
the effect achieved by maximizing bone contact through customization. In the authors’ opinion, in-vivo behaviour of the
personalized implant should lead to substantial increase in life service;
- Since the fatigue limit for both implants was much above the minimum allowed by ISO 7602: 2010, it can be concluded
that both models can be safely used in the medical practice.
As well, for these tests similar relevant studies were not found. A study to determine the fatigue limit [15] failed by
sample detaching from the embedding system and a simulation of fatigue behaviour of the femoral stem without
specifying the fatigue limit value is presented by Jui-Pin Hung et al. in [20].
<H2>4.3. The study of the crystallographic structure
In this study, three important conclusions resulted:
- There are more pores in the rolled bar used as raw material for dental implants made by chip removing processing than
in the implants made by laser sintering as additive manufacturing, which showed a fully dense structure (only isolated
very small pores). This could be explained by the difference between the two technologies, finer micro-structure being
obtained using laser sintering for additive manufacturing;
- Crystallographic structure of the implant made using laser sintering revealed a reinforced micro-structure of the
material, with strings along different directions for each layer, as laser changes the scrolling direction every layer. This
means that the tensile resistance of the implant could be greater than the resistance of the rolled bars of titanium.
- Having less pores then the rolled bar, it is expected that the sintered material is tougher because pores could represent
mechanical tensions concentrators, favouring the formation of micro-cracks that has a direct influence upon material
toughness, so the implant could be able to absorb more mechanical energy without break.
As a future research direction, toughness and brittleness of the titanium samples obtained by rolling and laser sintering
should be compared.
<H1>5. Conclusions
Research carried out by static tests have shown that realization of the fenestrations that mimics natural bone structure
conferred to the implants an elastic behaviour proven by the nonlinear appearance of the characteristic curves. In the
implant with large fenestrations the curve revealed a pronounced nonlinear aspect, recalling the rubber elastic behaviour,
showing large strains at smaller forces, displacements that decreased with the increase of the loading force. Increasing
elasticity is of great importance because an implant elasticity close to that of the natural bone contributes to reducing the
stress shielding which is one of the major causes of hip implant rejection [21]. Fenestrations also contribute to implant
osseointegration ensuring better anchoring and greater mechanical stability.
Dynamic tests allowed determination of the fatigue limit and showed that large fenestrations conferred the desired
elasticity to the implant, and because the fatigue limit for both implants was much above the minimum value allowed by
actual standard, it can be concluded that both models can be used safely in medical practice.
Future work will be focused on studies in order to confirm the better anchoring and better mechanical stability due to
bone ingrowth in the fenestrated architecture, which at this moment can only be intuited and predicted by computer
simulations [22].
<H1>6. Funding sources
The research was funded by a Romanian research contract PNCDI II, Programme 4: ''Partnerships in priority areas'', 2008
Competition: Advanced computer system based on medical imaging for producing personalized implants dedicated to
total hip arthroplasty, Acronym: SABIMAS, Contract No. 12-107 / 01.10.2008.
<REF>References
<BIBL>
[1] Darrell L.,;1; Moulton Ronald W. Lindsey: Proximal Femur Size and Geometry in Cementless Total Hip Arthroplasty
Patients, http://f1000research.com/articles/4-161/v1, (last <C>accessed 27.04.2016).</C>
[2] M.I.Z. Ridzwan, Solehuddin Shuib, A.Y. Hassan, A.A. Shokri and M.N. Mohamad Ibrahim,;1; 2007. Problem of
Stress Shielding and Improvement to the Hip Implant Designs: A Review. Journal of Medical Sciences, 7: 460-467.
[3] Orban H., Bunea D., Moldovan L., Antoniac I., Gheorghiu D., Semenescu A.;1;: Componentă femurală pentru
proteză totală de șold, Brevet de invenție RO 117890 B, Oficiul de stat pentru invenții și mărci, 30 Sept. 2002.
[4] Lawrence E. Murr, Sara M. Gaytan, Edwin Martinez, Frank Medina, Ryan B. Wicker:;1; Next Generation
Orthopaedic Implants by Additive Manufacturing Using Electron Beam Melting, International Journal of Biomaterials,
Volume 2012 (2012), <C>Article ID 245727, 14 pages http://dx.doi.org/10.1155/2012/245727 (last <C>accessed
07.02.2017)</C></C>
[5] Arabnejad S., Johnston B., Tanzer M., Pasini D.:;1; Fully porous 3D printed titanium femoral stem to reduce stress-
shielding following total hip arthroplasty, J Orthop Res. 2016, Sep 24, <DOI> 10.1002/jor.23445.</DOI>

4
[6] Harrysson O. L. A., Cansizoglu O., Marcellin-Little D. J., Cormier D. R., West H. A.:;1; Direct metal fabrication of
titanium implants with tailored materials and mechanical properties using electron beam melting technology. Materials
Science and Engineering C. 2008;28(3):366–373. <DOI> 10.1016/j.msec.2007.04.022</DOI>
[7] Hrabe N. W., Heinl P., Flinn B., Körner C., Bordia R. K.:;1; Compression-compression fatigue of selective electron
beam melted cellular titanium (Ti-6Al-4V), J Biomed Mater Res B Appl Biomater. 2011 Nov; 99(2):313-20. <DOI>
10.1002/jbm.b.31901.</DOI>
[8] MIMICS: Software & Services for Biomedical Engineering, http://biomedical.materialise.com/mimics, (last
<C>accessed 28.03.2015)</C>
[9] SolidWorks: 3D CAD solutions, http://www.solidworks.com/sw/products/3d-cad/packages.htm (last
<C>accessed 15.06.2016)</C>
[10] EOS e-Manufacturing Solutions: Industrial 3D Printing, http://www.eos.info/en (last <C>accessed
23.05.2016)</C>
[11] ***: ISO 7206-4:2002: Implants for surgery - Partial and total hip prostheses, Part 4-Determination of
endurance properties of stemmed femoral components.
[12] ***: ISO 7206-8:2002: Implants for surgery - Partial and total hip prostheses, Part 8- Endurance
performances of stemmed femoral components.
[13] A. Pacioga:;1; Research On the Development of a New Conceptual Base in the Design and Manufacture of
Hip Prostheses through Customization, PhD Thesis, POLITEHNICA University of Bucharest, 2011.
[14] ***: Tinius Olsen Testing Machine Company, https://www.tiniusolsen.com/(last <C>accessed
08.07.2016)</C>
[15] ***: Elements of Metallurgy and Engineering Alloys, Chap. 14: Fatigue, ASM International® <C>ISBN: 978-0-
87170-867-0, http://www.asminternational.org/documents/10192/1849770/05224G_Chapter14. pdf (last <C>accessed
08.07.2016)</C></C>
[16] ***: Inverted microscope ECLIPSE MA100/MA100L,
http://www.nikonmetrology.com/en_EU/Products/Microscope-Systems/Inverted-Microscopes/Eclipse-MA100-
MA100L (last <C>accessed 01.05.2016)</C>
[17] ***: Imaging Software NIS-Elements F Ver4.30.01 for 32bit edition, http://www.nikon.com/products/microscope-
solutions/support/download/software/imgsfw/nis-f_v4300132. htm (last <C>accessed 29.06.2016)</C>
[18] S.M. Croitoru, A. Pacioga:;1; Testing Titanium Implants, Proceedings in Manufacturing Systems, Vol. 7,
Iss. 1, 2012 / 1-10
[19] Ilaria Campioni, Gianluca Notarangelo,;1; Ugo Andreaus, Claudia Giacomozzi: Hip Prostheses Computational
Modeling: FEM Simulations Integrated with Fatigue Mechanical Tests, Biomedical Imaging and Computational
Modeling in Biomechanics, Edition: 2013,<C> Chapter: 5, Publisher: <PN>Springer</PN>, <PL>Netherlands</PL>,
pp.81-108</C>
[20] Jui-Pin Hung, Jian-Horng Chen, Hsiu-Lu Chiang,;1; James Shih-Shyn Wu: Computer simulation on fatigue
behavior of cemented hip prostheses: a physiological model, Computer Methods and Programs in Biomedicine (2004)
76, 103-11
[21] W.D. Bugbee, W.J. Culpepper 2nd, C.A. Engh Jr, C.A. Engh Sr.:;1; Long-term clinical consequences of stress-
shielding after total hip arthroplasty without cement, J Bone Joint Surg Am. 1997 Jul;79(7):1007-12.
[22] Andriy Andreykiv et. al.;1;: Simulation Of Bone Ingrowth, Printed by PrintPartners Ipskamp,2006, <C>ISBN-10:
90-9021020-2, ISBN-13: 978-90-9021020-9, pp. 77-88.</C>

</BIBL>
<Figure>Figure 1: Implant manufacturing

<Figure>Figure 2: Orientation and fixation in gripping device

<Figure>Figure 3: Implant fixture on H10KT machine

<Figure>Figure 4: Locati method

<Figure>Figure 5: Dynamic testing/thermostatic bath

<Figure>Figure 6: Characteristic curves

<Figure>Figure 7: Graphs of the Locati test - personalized implants

5
 <Figure>Figure 8: Crystallographic structure of Ti-6Al-4V (800x)
Cur. Minimum Maximum Number of [18]
no. force [N] force [N] cycles
1 230 2,300 5,000,000
2 280 2,800 1,000,000
<Table>Table 1: Parameters of Locati test
3 330 3,300 1,000,000
4 380 3,800 1,000,000
5 430 4,300 1,000,000 <Table>Table 2: Vertical displacements for static testing
6 480 4,800 1,000,000
Average
Cur. Sample Loading force Displacement
Implant model displacement
no. number [N] [mm]
[mm]
1 I 2.431
Large
2 II 2.112 2.280
fenestrations
3 III 2.297
5 I 2700 0.981
Small
6 II 0.910 0.931
fenestrations
7 III 0.903
9 I Commercial 1.562 1.562

<Table>Table 3: Locati test results

Results
Number of Estimated Test Total duration
Loading Fatigue limit
Implant cycles fatigue limit duration
force F0 [N] FD [N]
model Δn (NL) FDL [N] [h]
[h] [days]
6
2,300 5·10 138.9
2,800 106 27.8
3,300 106 27.8
Commercial 3,800 106 4,832 4,590 27.8 279.7 11.6
4,300 106 27.8
4,800 106 27.8
5,300 63,428 1.8
2,300 5·106 138.9
Large 2,800 106 27.8
3,554 3,376 208.6 8.7
fenestrations 3,300 106 27.8
3,800 508,211 14.1
2,300 5·106 138.9
2,800 106 27.8
3,300 106 27.8
Small
3,800 106 4,928 4,682 27.8 284.97 11.9
fenestrations
4,300 106 27.8
4,800 106 27.8
5,300 256,032 7.2

TDENDOFDOCTD