diff --git a/solver/src/model/lang/linear.rs b/solver/src/model/lang/linear.rs
index fb967890..28e2adfd 100644
--- a/solver/src/model/lang/linear.rs
+++ b/solver/src/model/lang/linear.rs
@@ -338,8 +338,8 @@ impl<T: Into<LinearSum>> std::ops::Add<T> for LinearSum {
     type Output = LinearSum;
 
     fn add(self, rhs: T) -> Self::Output {
-        let mut new = self;
-        new += rhs.into();
+        let mut new = self.clone();
+        new += rhs;
         new
     }
 }
@@ -352,6 +352,16 @@ impl<T: Into<LinearSum>> std::ops::Sub<T> for LinearSum {
     }
 }
 
+impl<T: Into<i32>> std::ops::Mul<T> for LinearSum {
+    type Output = LinearSum;
+
+    fn mul(self, rhs: T) -> Self::Output {
+        let mut new = self.clone();
+        new *= rhs;
+        new
+    }
+}
+
 impl<T: Into<LinearSum>> std::ops::AddAssign<T> for LinearSum {
     fn add_assign(&mut self, rhs: T) {
         let rhs: LinearSum = rhs.into();
@@ -369,6 +379,16 @@ impl<T: Into<LinearSum>> std::ops::SubAssign<T> for LinearSum {
     }
 }
 
+impl<T: Into<i32>> std::ops::MulAssign<T> for LinearSum {
+    fn mul_assign(&mut self, rhs: T) {
+        let rhs = rhs.into();
+        self.constant *= rhs;
+        for term in self.terms.iter_mut() {
+            term.factor *= rhs;
+        }
+    }
+}
+
 impl std::ops::Neg for LinearSum {
     type Output = LinearSum;
 
@@ -1041,6 +1061,16 @@ mod tests {
         assert_eq!(result.terms, vec![]);
     }
 
+    #[test]
+    fn test_sum_mul() {
+        let v = IVar::new(VarRef::from_u32(5));
+        let s = LinearSum::of(vec![FAtom::new(IAtom::new(v, 5), 28)]);
+        let result = (s * 3).simplify();
+        assert_eq!(result.constant, 15);
+        assert_eq!(result.denom, 28);
+        assert_eq!(result.terms, vec![LinearTerm::new(3, v, 28)]);
+    }
+
     #[test]
     fn test_sum_sub() {
         let s1 = LinearSum::of(vec![FAtom::new(5.into(), 28)]);
@@ -1075,6 +1105,18 @@ mod tests {
         assert_eq!(result.terms, vec![]);
     }
 
+    #[test]
+    fn test_sum_mul_assign() {
+        let v = IVar::new(VarRef::from_u32(5));
+        let s = LinearSum::of(vec![FAtom::new(IAtom::new(v, 5), 28)]);
+        let mut result = s.clone();
+        result *= 3;
+        let result = result.simplify();
+        assert_eq!(result.constant, 15);
+        assert_eq!(result.denom, 28);
+        assert_eq!(result.terms, vec![LinearTerm::new(3, v, 28)]);
+    }
+
     #[test]
     fn test_sum_neg() {
         let s1 = LinearSum::of(vec![FAtom::new(5.into(), 28)]);