Evaluasi Empiris dan Pemodelan Prediktif Durabilitas Warm Mix Asphalt terhadap Siklus Beku - Cair pada Iklim Dataran Tinggi

Abstract

Isu durabilitas perkerasan jalan pada kondisi iklim ekstrem menjadi krusial seiring meningkatnya frekuensi siklus beku–cair di dataran tinggi, sementara bukti empirik mengenai respons Warm Mix Asphalt (WMA) terhadap fenomena tersebut masih terbatas. Penelitian ini bertujuan untuk mengevaluasi secara kuantitatif pengaruh siklus beku–cair terhadap durabilitas WMA dan membandingkannya dengan Hot Mix Asphalt (HMA). Studi ini dilaksanakan dengan pendekatan eksperimen laboratorium yang divalidasi data lapangan pada 1–2 lokasi dataran tinggi; sampel lab (n=40; WMA=20, HMA=20) dan core lapangan (n=24) were prepared and tested for physical and mechanical properties (VTM/VMA, ITS, Marshall, dynamic modulus, rutting, fatigue), and predictive models were developed using regression and Random Forest. Hasil menunjukkan bahwa setelah 50 siklus WMA mengalami peningkatan porositas +12.0% (CI95%: 9.1–14.9%) dan VTM +34% (3.8→5.1%), sedangkan HMA masing-masing +6.0% dan +20%; ITS WMA turun 28.0% (600→432 kPa) versus HMA 18.0% (640→525 kPa) (mixed-ANOVA: siklus F(3,32)=45.6, p<0.001; interaksi p=0.009). Laju degradasi ITS per 10 siklus adalah β=−0.056 (WMA) dan β=−0.036 (HMA); median cycles-to-failure 42 (WMA) vs 58 (HMA). Model Random Forest mencapai R²validasi=0.81. Hasil penelitian ini menunjukkan bahwa paparan freeze–thaw secara kuantitatif mempercepat penurunan kinerja WMA dibanding HMA; implikasinya adalah perlunya penyesuaian desain mix, inspeksi lebih sering, dan kalibrasi faktor lingkungan dalam kerangka Mechanistic-Empirical untuk perencanaan perkerasan di dataran tinggi.

The durability of pavement materials under extreme climatic exposure is a pressing concern; however, empirical evidence on Warm Mix Asphalt (WMA) performance under freeze–thaw cycles in highland environments is scarce. This study aims to assess quantitatively the influence of freeze–thaw cycles on WMA durability and to compare its performance with Hot Mix Asphalt (HMA). A laboratory experiment with field validation was conducted at 1–2 highland sites; laboratory samples (n=40; WMA=20, HMA=20) and field cores (n=24) were tested for physical and mechanical indicators (VTM/VMA, ITS, Marshall stability, dynamic modulus, rutting, fatigue), and predictive models (regression and Random Forest) were developed. Results showed that after 50 cycles WMA exhibited a mean porosity increase of +12.0% (95% CI: 9.1–14.9%) and VTM rise of +34% (3.8→5.1%), while HMA increased +6.0% and +20% respectively; ITS decreased 28.0% for WMA (600→432 kPa) versus 18.0% for HMA (640→525 kPa) (mixed-ANOVA: cycles F(3,32)=45.6, p<0.001; interaction p=0.009). Degradation rates per 10 cycles were β=−0.056 (WMA) and β=−0.036 (HMA); median cycles-to-failure were 42 (WMA) and 58 (HMA). The Random Forest model achieved an R² validation of 0.81. These findings indicate that freeze–thaw exposure substantially accelerates WMA degradation relative to HMA. The implications include the need for mix design adjustments, more frequent inspection regimes, and the incorporation of empirically calibrated environmental damage factors into Mechanistic-Empirical pavement design frameworks.